Spring 2009, Vol. 7, No.

Qualitative &
Multi-Method Newsletter of the
American Political Science Association

Research Organized Section for Qualitative and

Multi-Method Research

Contents Letter from the Editor

Symposium: Teaching Interpretive Methods
Introduction
Peregrine Schwartz-Shea ........................................................2
Making Room for Intepretivism? A Pragmatic Approach
Robert Adcock ..........................................................................3
Undoing the Opposition Between Theory and Methods
Emily Hauptmann .....................................................................8
Introducing Rigor to the Teaching of Interpretive Methods
Jan Kubik .................................................................................11
Teaching Interpretive Methods in Political Science: The
Challenges of Recognition and Legitimacy
Peregrine Schwartz-Shea ......................................................18
Teaching Gender and Politics: Feminist Methods in
Political Science
Mona Lena Krook ..................................................................24
Symposium: Conceptualizing and Measuring
Ethnic Identity
Measuring the State’s Institutionalization of Ethnic
Categories across Time and Space
Evan S. Lieberman and Prerna Singh ...................................29
Design Measures of Ethnic Identity: The Problems of
Overlap and Incompleteness
Kanchan Chandra ..................................................................36
Measuring Identity Content and Comparing Social
Identities
Yoshiko Herrera, Rose McDermott, Alastair Iain Johnston,
and Rawi Abdelel ...................................................................41
Book/Article Notes ........................................................46
Announcements ...................................................................48
APSA-QMMR Section Officers
President: John Gerring, Boston University
President-Elect: Colin Elman, Syracuse University
Vice President: Margaret Keck, Johns Hopkins University
Secretary-Treasurer: Colin Elman, Syracuse University
Newsletter Editor: Gary Goertz, University of Arizona
Division Chair: Rudra Sil, University of Pennsylvania
Executive Committee: Rose McDermott, Brown University
Peregrine Schwartz-Shea, University of Utah
Page Fortna, Columbia University
Dan Carpenter, Harvard University
Gary Goertz
University of Arizona
ggoertz@u.arizona.edu
As has become customary in the Spring issue of the news-
letter, we list the section’s panels for APSA 2009. Rudy Sil
(rudysil@sas.upenn.edu) has done a great job in managing
and organizing the panels. We have about 25 panels this year.
As usual there is considerable diversity in the panels, which
reflects the diversity of the section membership. Panels may
change between now and APSA so you should check the APSA
website for the final program. This is also the occasion for me
to correct a mistake in the last issue and thank Craig Thomas
for his work in organizing the panels for APSA 2008.
I also encourage members to attend the business meeting
and reception at APSA. The business meeting is usually on
Thursday evening. It is a good chance to hear about the sec-
tion; it is an even better occasion to have some interesting
chats during the reception. Check the section website (http://
www.maxwell.syr.edu/moynihan/programs/cqrm/section.html)
for information about the business meeting, receptions, work-
shops, and other events.
One of the goals of the newsletter is to provide useful
information and readings for the teaching of qualitative and
mixed methods. The symposium on “Teaching Interpretive
Methods” provides some ideas and reflections on how one
can include interpretive methods in a qualitative methods class.
Some discussion of interpretive methods has become stan-
dard practice in qualitative methods classes so this sympo-
sium should be of use to many instructors.
Mona Krook reviews about 50 syllabi for gender and poli-
tics classes, with particular attention to the methodology used
in the readings for such classes. The diversity of the method-
ologies represented in these courses may surprise some read-
ers.
As someone centrally interested in concepts and mea-
surement, when I saw the 2008 APSA panel on “New Ap-
proaches to the Measurement of Ethnicity: Identities, Institu-
tions, and Power,” I naturally thought this would make a good
symposium. I attended the panel and got some useful ideas for
my own seminars and workshops on concepts and measure-
ment. There has been a tremendous amount of work lately on
both the data on ethnic groups and the measurement of ethnic
group or identity variables. Ethnic identity variables are cru-
Qualitative & Multi-Method Research, Spring 2009
cial to quite a few research agendas, from economic growth to
civil war. At the same time, many feel that the ELF index and the
data used to construct it are very flawed. The Lieberman and
Singh, and Chandra articles show that one needs to include
political factors such as political parties and the political insti-
tutionalization of ethnic identities in the the conception of
ethnic identity and then in the collection of the data (as op-
posed to completely apolitical and individual-level ELF mea-
sure). The Herrera et al. piece surveys some of the core issues
involved in measuring identity.
As reported in the last issue, the Institute for Qualitative
and Multi-Method Research will be held in early June at Syra-
Symposium: Teaching Interpretive Methods
Introduction
Peregrine Schwartz-Shea
University of Utah
psshea@poli-sci.utah.edu
Even if one does not teach interpretive methods in par-
ticular, all those who teach methods can learn something from
these four symposium contributions as the essays provide grist
for reflection on pedagogical strategies and goals.1
For those readers specifically considering how to include
interpretive methods in their departmental curricula, these es-
says recount the challenges faced by professors currently
teaching such methods in the discipline. Chief among these
challenges is finding space in the curricula; each of these in-
structors includes interpretive methods as part of a broader
course, either research design (Adcock, Kubik, Schwartz-Shea)
or an advanced graduate seminar or an undergraduate Scope
and Methods course (Hauptmann).2 On the positive side, this
strategy means the opportunity to teach students who might
not otherwise take a stand-alone course (especially effective if
the course is in some way required3). On the negative side, the
exposure will perforce be quite limited, which raises the stakes
in the choice of topics and readings (as Adcock’s analysis
makes clear). Yet the fact that these inroads are occurring is
concrete evidence of the impact of the last wave of disciplinary
debates over methodological pluralism.
In this context, Emily Hauptmann’s contribution, “Undo-
ing the Opposition Between Theory and Methods,” provides a
fitting history lesson on the origins of disciplinary assump-
tions about “methods” and “methods teaching:” (a) that meth-
ods training necessarily entails an uncritical perspective on
political life; (b) that political theorists have no methods; and
(c) that theorists should not be involved in methods teaching.
When political theorists contribute to their departmental meth-
ods curricula at either the graduate or undergraduate level,
these assumptions can be surfaced and revisited—a healthy
development, even if, as Hauptman puts it, she realizes “how
differently many of my colleagues think about [methods]” com-
pared with her own perspective.
2
cuse University (http://www.maxwell.syr.edu/moynihan/pro
grams/cqrm/institute.html). Along with a new venue comes a
new structure. The Institute is now being organized into two-
day modules to allow for more in-depth teaching of various
topics. Due to the increasing amount of material in the area of
qualitative and mixed methods, there will be two concurrent
modules that students can choose from, along with some com-
mon sessions for everyone. I am excited about the new loca-
tion and the new structure for the Institute and I think it repre-
sents a new stage in the evolution of the teaching of qualita-
tive and mixed methods within political science.
Recognizing interpretive research and the methods in in-
terpretive research is also a theme common to the other es-
says. Both Kubik and Schwartz-Shea remark on the taken-for-
granted status of interpretive methodologies in other disci-
plines; their discussion and citations also demonstrate that
such methods are already part of political science if attention
is turned to particular parts of the discipline (e.g., legal studies,
feminism) or to substantive topics in the subfields (e.g., politi-
cal legitimacy, foreign policy, culture, identity, bureaucratic ex-
perience). Graduate students’ critical abilities are sharpened
when they are better able to identify and assess the methods
used in the interpretive research they encounter during their
coursework and, subsequently, throughout their careers.
How students react to interpretive readings and ideas is
carefully parsed in Robert Adcock’s essay, “Making Room for
Interpretivism: A Pragmatic Approach.” He reports that stu-
dents’ most positive responses are to those readings that dem-
onstrate interpretive contributions to explanation and field re-
search—whereas interpretivism’s ontological and epistemo-
logical stance and encouragement of reflexivity sometimes led
to “rejection, rather than recognition, of the claim of [interpre-
tive] views to be potential philosophical foundations for em-
pirical research in the political science.” In her essay, “Teach-
ing Interpretive Methods in Political Science: The Challenges
of Recognition and Legitimacy,” Peregrine Schwartz-Shea
shares a similar concern with student reaction to the possibil-
ity of interpretive research. Using an assignment that has stu-
dents design quantitative-behavioral, comparative case study,
and interpretive approaches to the same research topic pro-
duces, she argues, student recognition and appreciation for
their “methodological others”—those who may chose to ap-
proach topics in ways different from their own inclinations and
specialties.
In a manner that echoes Adcock’s approach, Jan Kubik’s
essay, “Introducing Rigor to Teaching Interpretive Methods,”
succinctly connects the strengths of interpretive methods and
methodologies to the concerns not only of Weber, but also of
rational choice sociologist James Coleman, survey researcher
Laura Stoker, and game-theoretic modeler Barry O’Neill. In this
way, students who most relate to these latter approaches can
find initial entrée to interpretive perspectives.

In sum, the challenges of teaching interpretive methods
have been met by these professors with a mixture of pragma-
tism, reflexivity, and innovation. And it can be argued that
their efforts enrich their departments. Just as theorists of de-
liberative democracy argue that citizens can learn the most by
listening intently to those with whom they disagree, so, too,
inclusion of interpretive methods in graduate methods cur-
ricula can contribute to the vitality of a department’s research
life—engendering debate in its graduate seminars and depart-
mental colloquia. At a minimum students can learn that there is
no need to fear their “methodological others.”
Notes
1
Thanks to the QMMR section for originally sponsoring this
roundtable at the 2008 APSA conference in Boston and to newsletter
editor Gary Goertz for encouraging the contributors to formalize
their remarks for this symposium. Contributors’ syllabi are available
by emailing the authors or at the CQRM website hosted at the Max-
well School: http://www.maxwell.syr.edu/moynihan/programs/cqrm/
syllabi.html.
2
Although Kubik reports working on a new stand-alone course on
interpretivism, it is likely that such courses are still relatively rare in
the discipline. For an exception, see Lisa Weeden’s quarter-length
course, Interpretive Methods in the Social Sciences, at the Univer-
sity of Chicago. As described in the course catalog, “This course is
designed to provide students with an introduction to interpretive
methods in the social sciences. Students will learn to ‘read’ texts and
images while also becoming familiar with contemporary thinking about
interpretation, narrative, ethnography, and social construction. Among
the methods we shall explore are: semiotics, hermeneutics, ordinary
language theory, and discourse analysis.” Often, stand-alone courses
first become available under special topics numbers; see, e.g., a 2009
offering by Ido Oren at the University of Florida, Interpretive Ap-
proaches to Political Science. Oren’s syllabus is available through the
CQRM website address in note 1.
3
For example, as Robert Adcock explains, graduate students at his
institution must take either an advanced statistics course or the re-
search design course he describes in his essay.
Making Room for Interpretivism?
A Pragmatic Approach
Robert Adcock
George Washington University
adcockr@gwu.edu
Interpretivist scholars have carefully documented the mini-
mal, at best, presence that interpretive philosophical perspec-
tives and empirical methods have had in political science meth-
ods texts and curricula (Yanow and Schwartz-Shea 2002;
Schwartz-Shea 2003). What are we to make of this absence? Is
there a problem to be rectified here? Or an allocation decision
justifiable in light of limited pedagogical time and resources?
Given the profusion of philosophical perspectives and meth-
ods for accessing, generating, and analyzing data found in the
social sciences as a whole, some absences are unavoidable in
any single discipline. Thus, the fact of the relative absence of
interpretivism in the methods training of political science gradu-
Qualitative & Multi-Method Research, Spring 2009
ate students cannot alone support arguments for (or against)
giving it more room.1
I expect readers of this newsletter differ, perhaps even
strongly, regarding the imperative (or lack thereof) to expand
exposure to interpretivism. Moreover, these differences are, I
suspect, tied up with alternative assumptions about how gradu-
ate students would respond if interpretivism were given more
room in methods training. One potential assumption is that
greater exposure would lead more students to use interpretive
perspectives or methods in their research. Alternatively, in-
creased exposure might be expected to produce more confu-
sion than conversions. The “conversion” assumption is prob-
ably more common among advocates of greater attention to
interpretivism, and the “confusion” assumption among skep-
tics.2 A third possibility would be a “recognition” assumption
that increasing exposure may make graduate students more
likely to see interpretive research as falling within the disci-
plinary parameters of political science, without necessarily
making them more likely to undertake it themselves. The latter
assumption initially motivated me to experiment with giving
interpretivism some room in my methods teaching. But, as I
explore below, my experiences have subsequently led me to
rethink that initial assumption.
My goal in this piece is to promote treating these kinds of
assumptions not as the sacred hopes (or fears) of sects fight-
ing for the souls of students, but as tentative hypotheses.
Evaluating them requires introducing some students to inter-
pretive perspectives and methods, and reflecting upon the re-
sults. Methods instructors comfortable enough with interpretiv-
ism to give it room in their teaching are, however, usually fa-
vorably predisposed toward it. To counteract confirmation bias,
those of us who make room for interpretivism must be espe-
cially attentive to the possibility of minimal or even negative
outcomes. A self-critical frankness is essential if our reflec-
tions on our experiences are to be received by disciplinary
colleagues as imbued not with methodological partisanship,
but with pedagogical pragmatism. A pragmatic approach to-
ward making room for interpretivism must reflexively seek out,
and critically adapt in light of, the practical lessons of peda-
gogical experiments, whether those experiments turn out as
initially hoped or not.
What Kind of Methods Course do I Teach to Whom?
There is no single recipe for giving interpretivism more
room in methods training. Efforts could involve anything from
adding an interpretive reading or two to an existing syllabus,
to designing a full course, or even a multi-course curriculum,
exploring interpretivism in its rich variety. My own effort has
been limited. I have added interpretive readings to an existing
graduate methods course while taking key parameters of the
course as fixed. This newsletter’s sophisticated readers will be
better judges than I of which aspects of my effort, and the
practical lessons I draw from it, might transfer to their own
pedagogical contexts. But for readers to make such judgments
it is necessary that I spell out some details regarding the kind
of course I have been teaching, in what broader curricular set-
ting, and to whom.
3
Qualitative & Multi-Method Research, Spring 2009
I teach a semester-long graduate course entitled “System-
atic Inquiry and Research Design” (PSc 209) to students who
are nearly all pursuing a political science PhD. PSc 209 com-
bines a short philosophy of science component with more
extended readings on research design, and culminates with
students writing and critiquing draft research proposals. It is
part of my department’s methods sequence, which begins with
a course introducing statistical thinking and tools. All PhD
students take that course (or have taken its equivalent else-
where), and they then complete their methods requirement ei-
ther with an additional semester of statistical training, or by
taking PSc 209.
The curricular setting of PSc 209 has two consequences
for the abilities and interests of students. First, they have (or
are acquiring) some familiarity with basic descriptive and infer-
ential statistics. This has notable payoffs. For example, it helps
students to engage with nuanced arguments about the con-
ceptions of causation statistical techniques draw upon (on
this topic I assign Abbott 1998 and Goldthorpe 2001). Second,
students often take PSc 209 in lieu of a further statistical course
if they plan to pursue primarily qualitative dissertation research.
Many students thus come to the class looking quite specifi-
cally for guidance in qualitative methods. Some students are,
however, designing quantitative research, and many are inter-
ested in multi-method approaches. In terms of subfield distri-
bution, international relations is the major field of the largest
number of students, followed by comparative politics, and lastly
American politics.
What Aspects of Interpretivism to Include?
My effort to make room for interpretivism is only one, and
not the most important, pedagogical goal shaping the content
of my PSc 209 syllabus. I thus have, at most, a few weeks of
readings to play with, and can introduce only a taste of the
philosophical perspectives and empirical methods of access-
ing, generating, and analyzing data that might be labeled inter-
pretive. One response to these limits would be to focus the
readings and explore in some depth a specific way of framing
interpretivism (for example, the framing in Adcock 2003). But I
opted instead to experiment with a more diverse set of read-
ings.
The starting point for my pragmatic approach was to scat-
ter readings reflecting various ways of approaching interpret-
ivism throughout the semester. I did not use any kind of sam-
pling frame in selecting readings, but in retrospect they may be
summed up as offering at least two readings for each of four
entry points to interpretivism:
(1) Interpretivism as a general epistemological and/or onto-
logical stance;
(2) Interpretivism as a stance specifically centered on ques-
tions of explanation in social science;
(3) Interpretivism as research emphasizing “reflexivity”;
(4) Interpretivism as field research methods that seek un-
derstanding of others through intensive interaction in day-
to-day settings.
I approached my readings as a series of practical experi-
4
ments—something akin to canaries in a goldmine—to explore
student reactions to varied aspects of interpretivism. I sought
to gauge reactions both during seminar discussions, and via
end-of-semester ratings of readings.3 My interpretation of
these reactions then informed syllabus revisions for my next
iteration teaching PSc 209: I trimmed two of the four entry
points listed above (1 and 3), while expanding attention to the
other two (2 and 4).4
In the four sections below, I discuss, for each entry point
in turn, the readings I assigned, student reactions, and my
subsequent syllabus revisions. It would be disingenuous to
present this process as if I were a dispassionate experimenter
coolly observing reactions. I went into the course with rather
well-developed priors regarding the strengths or weaknesses
of alternative formulations of interpretivism (Adcock 2003). I
also had assumptions regarding how students would respond
to various aspects of interpretive philosophy and methods. If
my subsequent syllabus revisions are one register of the re-
sults of my pedagogical experimentation, another is the updat-
ing of my own beliefs. I include below some commentary on
this more personal intellectual dimension because pragmatism
involves more than practical experimenting and adaptation; it
also involves being reflexive about the role that the subjectiv-
ity of the scientist/scholar plays in these processes.
Entry Point #1: Epistemology and Ontology
Contemporary interpretivists claim the status of an au-
tonomous third position within a tripartite division of the ter-
rain of methodology and methods that includes also quantita-
tive and qualitative positions (Yanow and Schwartz-Shea 2006:
xv–xiv). The claim relies, first and foremost, on the argument
that interpretivists share a distinctive epistemological and on-
tological stance which sets them apart from a diffuse “positiv-
ism,” seen as the philosophical common ground of quantita-
tive, and also most qualitative, political scientists. This argu-
ment was (and remains) my least favorite entry point to
interpretivism (see Adcock 2009).Yet philosophical abstinence
did not appear plausible either, because leading currents in
recent methodological conversations—such as wide embrace
of the ideal of “shared standards” (Adcock and Collier 2001;
Brady and Collier 2004) among quantitative and qualitative
scholars—do rely on epistemological premises questioned by
many interpretivists.
My initial syllabus included two readings chosen to spot-
light these issues. I selected the sixth edition of Neuman’s
Social Research Methods (2006) as a text largely because of
its material on epistemology and ontology. Neuman identifies
three major longstanding philosophical traditions in social sci-
ence—positivist, interpretive, and critical—and surveys their
positions across a range of epistemological and ontological
issues. He also introduces feminism and postmodernism as
more recently developed stances that may inform empirical
social science. Neuman’s discussion is a textbook example of
the pros and cons of textbooks. It is accessible and concisely
summed up in a helpful table (2006: 105). But to pull this off
nuances are eliminated. As a more advanced reading a little
later in the class, I assigned “What would an adequate philoso-

phy of social science look like?” by Fay and Moon (1977).
They contrast traditions of “naturalism” and “humanism,” and
then argue that a social science capable of “critique” must
transcend this dichotomy. Fay and Moon here use slightly
different terminology to engage the same three traditions as
Neuman, but they advance a philosophical argument between
traditions, rather than a textbook survey.
I assumed that some of my students would be attracted to
one or another of these anti-positivist philosophical traditions,
and some confused by them. I hoped, in turn, that the attracted
students would spark a class discussion which, by differenti-
ating alternative anti-positivist stances and debating them,
would help clarify philosophical issues for them and their class-
mates.5 But I was entirely mistaken. In the class sessions for
which the Neumann and Fay and Moon readings were as-
signed, students did not raise the alternatives to positivism
themselves, and when I pushed this material into one discus-
sion, they reacted by identifying with positivism. Indeed, the
main effect of my interjection was to spur students to argue
that alternative positions fail to meet key norms of political
science, such as objectivity. In sum, rather than loosening
positivist tendencies in my students’ developing sense of dis-
ciplinary identity, presenting epistemological and ontological
challenges only provided an “other” which reinforced those
tendencies. Exposure led to rejection, rather than to recogni-
tion, of the claim of other views to be potential philosophical
foundations for empirical research in political science.
In light of this experience, I dropped this first entry point
to interpretivism from the next iteration of PSc 209. My deci-
sion was made easier by the discovery that other readings
were effective at getting students to debate shared standards.
The reading that worked best here was Mahoney and Goertz’s
“A Tale of Two Cultures” (2006), which highlights differences
in how quantitative and qualitative scholars tend to conceive
of explanation and causation, and then traces the ramifications
of these differences through a wide array of research norms
and practices. Mahoney and Goertz synthesized specific con-
trasts that were tangibly familiar to students from readings in
their substantive classes. The pedagogical take-home for me
was that exploring how any one standard—such as “causal
inference”—is actually pursued in familiar political science ex-
amples is a more effective spur to discussion of whether we
have (or should have) “shared standards” than staging a philo-
sophical “battle royale” at the level of epistemology and on-
tology.
Entry Point #2: What Makes for a Good Explanation?
My second entry point to interpretivism, while still philo-
sophical in character, was pitched at the level of a specific
question: what makes for a good explanation in social science?
I incorporated interpretive views on this question within a
several-week unit on explanation and causality. I devoted most
of one session to the classic debate regarding the relation
between understanding and explanation, assigning Charles
Taylor’s “Interpretation and the Sciences of Man” (1971), along-
side readings selected to engage subfield interests in Ameri-
can politics (Soss 2006), comparative politics (Schwartz 1984;
Qualitative & Multi-Method Research, Spring 2009
Kurzman 2004), and international relations (Wendt 1998).
This second entry point to interpretivism was the most
successful in raising sympathetic student engagement. In broad
outline, our class discussion tracked a path that I expected.
Students were receptive to the argument that gaining under-
standing is a core task of social science, so long as under-
standing was framed in Weberian fashion as an aid, rather than
alternative, to explanation. I was, however, surprised by stu-
dents’ reactions to two readings. First, put bluntly, they hated
Taylor’s famous essay. They found it too abstract, perhaps in
part because they lacked familiarity with its examples from po-
litical science debates of the 1960s. Second, the students were
excited by Wendt’s argument that there is more than one kind
of explanation. These reactions inverted my own priors, which
ranked Taylor as the best reading, and underrated Wendt’s
concept of “constitutive explanation” as a mode of explana-
tion distinguished from “causal explanation.”
In light of these reactions, I subsequently revised my
syllabus in three ways. First, I cut Taylor’s article since it just
does not speak effectively to my graduate students. Second,
I reevaluated my view of Wendt’s distinction between causal
and constitutive explanations. I decided that, rather than a
confusing novelty, it links up nicely with the distinction be-
tween causal and unification theories of explanation presented
in the philosophy of science text I assign (Godfrey-Smith 2003:
Chap. 13). It is, moreover, useful for unpacking the “interpretist”
approach of Schwartz (1984) as being explanatory in a specifi-
cally constitutive sense. Third, I added readings to explore
varieties of explanation far more fully. In particular, I chose
readings which, like Wendt, distinguish kinds of explanations,
but that draw distinctions along different lines: Roth (2004)
on “structural” vs. “situational” vs. “intentional” explana-
tions, and Abbott (2004) on “semantic” vs. “syntactic” vs.
“pragmatic” explanatory programs. Finally, to balance the split-
ting tendency of these readings, I added material from Elster’s
new Explaining Social Behavior arguing that all good social
science explanations share common features based on the
hypothetico-deductive method (2007: 15–20).
These syllabus shifts supplemented the classic issue I
began with—the relation of understanding to explanation in
social science—with questions about whether there are mul-
tiple varieties of explanation, and if so, whether there are inte-
grating standards relevant across them all. Raising these ad-
ditional questions remakes the old debate about understand-
ing as an aid or an alternative to explanation. The debate
becomes whether understanding is necessary to all species
of social science explanation (and thus offers a shared stan-
dard in light of which all explanations that do not incorporate
understanding fall short), or whether it is, instead, necessary
only for some kinds of explanation (and thus a distinguishing
feature of those kinds). My beliefs on these questions are
unsettled. But I hope teaching my methods class will con-
tinue to push along my thinking in the years ahead!
Entry Point #3: Reflexivity
It is common for interpretivists to emphasize the need for,
and benefits of, greater “reflexivity” in social science. Reflex-
5
Qualitative & Multi-Method Research, Spring 2009
ivity involves scholars reflecting upon their own knowledge- Entry Point #4: Intensive Interaction in Day-to-Day Settings
producing activities with the same tools and critical distance
My fourth entry point to interpretivism focused on gath-
they apply when analyzing the activities of others. This can
ering data through intensive in-person interaction with sub-
be done at an individual level, with researchers considering,
jects in their day-to-day settings. While such interaction is a
for example, how their personal identity characteristics influ-
core part of ethnography, I wanted to emphasize that it is a
ence their research activities and products. My initial PSc 209
broader data-gathering method. So in selecting readings I
syllabus included one reading chosen to introduce this per-
paired the anthropologist Clifford Geertz’s classic “Thick De-
sonalized reflexivity (Shehata 2006). Reflexivity might, how-
scription” essay (1973) with readings from former APSA presi-
ever, be pursued with more aggregate units of analysis: re-
dent Richard Fenno (1977, 1990) about his participant obser-
search programs, subfields, disciplines, or even social sci-
vation of members of Congress in their home districts. A sec-
ence writ large.
ond goal was to highlight the endeavor to grasp the perspec-
I selected three readings exemplifying reflexivity in its
tive of subjects who interpret and act in the world using con-
more sociological and historical forms. First, I assigned Oren’s
cepts different from those of the academics studying them. To
(2006) argument that major changes in the way that American
this end I assigned Schaffer (2006) on interviewing ordinary
political scientists conceptualize democracy have followed
citizens of other societies in a way designed to elicit details
shifts in America’s international relations, with “democracy”
and nuances of what a concept like “democracy” means in
reconceived to exclude nations that become America’s en-
their language and political context. I also assigned a further
emies and include its allies. A second reading emphasized the
essay by Geertz, “From the Native’s Point of View” (1979),
interplay between the ebb and flow of current events and
which reflects on the task of grasping the “experience-near”
intellectual shifts in political science, asking what this entails
concepts of subjects, but also contends that the ultimate goal
for the kinds of “progress” that are (and are not) possible in
is to relate those concepts to the “experience-distant” con-
our discipline (Dryzek 1986). The third reading (Osborne and
cepts of generalizing social science theory.
Rose 1999) looked at the dynamic between social science and
During class discussion of this day of readings, the dy-
society from the opposite direction, by asking whether social
namic of students identifying in contrast to an “other,” which
science can produce novel social phenomena. In particular,
had surprised me with regard to my first entry point, was again
the reading argued that the creation of sampling techniques
noticeable. The principal “other” here was Geertz and “eth-
remade not only social scientists’ knowledge of public opin-
nography” more generally, which my students saw as some-
ion, but ultimately the character of public opinion itself.
thing anthropologists, as opposed to political scientists, do.
I was personally excited about these three reflexivity read-
Some of this dynamic also developed in response to Fenno.
ings because they spotlight the history of social science, which
While recognizing his work as a classic in the American poli-
is one of my own research areas. However, the readings fell
tics subfield, my students identified his method of participant
decidedly flat. In each week for which one was assigned, stu-
observation as marginal to more recent developments in that
dents largely ignored it in favor of discussing other readings.
subfield. They then debated whether it could be a viable method
For example, I paired Oren’s article with Collier and Levitsky
today for young scholars of American politics anxious to be
(1997) work on conceptualizations of democracy, which ana-
seen as contributing to the cutting edge of the contemporary
lyzes a narrower period of political science usage and does
subfield. Hence, while I had chosen my Geertz and Fenno read-
not attend to the international context Oren emphasizes. It
ings as accessible pieces that might provide interesting mod-
was the latter piece students engaged with, however, because
els, students referred to them in our discussion first and fore-
it delved in more detail into specific conceptual developments
most as examples of what they believe they should not (or
in recent literature. When I flagged the contrasting approaches
cannot) do in their own research.
of the pieces, my students argued that Oren’s historical ap-
An illuminating contrast was, however, provided by the
proach did not offer the practical aid in clarifying conceptual
very positive reaction of my students to Schaffer (2006). With
confusions and choices they face in their own research, which
regard to endeavors to grasp the perspective of others, I ex-
they found in Collier and Levitsky’s work.
pected students to prefer Geertz’s “From the Native’s Point of
Taking account of such reactions, I have eliminated this
View” since it suggests that understanding local perspectives
third entry point to interpretivism from my syllabus. The sole
is not an end in itself, and that the social scientific goal in such
exception is Dryzek (1986), which I retain as part of a philoso-
work is to connect local perspectives to generalizing theory.
phy of science week on “progress.” A broader take-home of
But two points about the Schaffer piece won student apprecia-
this experience was to push me to think more carefully (and
tion. First, Schaffer presents his method as a mode of inter-
humbly!) about the status of reflexive historical analyses of
viewing. While in-depth field interviewing is a central part of
social science. While all methods courses encourage students
both ethnography and participant observation, about half of
to reflect upon the methods they use, I now incline to the view
my students singled out interviewing as something they plan
that specifically historical and sociological forms of reflexiv-
to do and want guidance in, even as they distanced them-
ity are two among many substantive research specializations,
selves from labels such as “ethnography” or “participant ob-
rather than methodological ideals for social scientists more
servation.” Second, Schaffer’s piece presents a lengthy ex-
generally to try to live up to.
cerpt of an actual interview to illustrate his differentiation of

6

types of questions and their role at different points in an inter-
view. It thus offered students a concrete sense of what in-
depth interviewing can involve, which turned out to be much
closer to what they were looking for than the more meta-reflec-
tions of Geertz’s essay.
In revisiting my syllabus in light of these reactions, I
dropped Geertz’s “Native’s Point of View” and turned inter-
viewing into a central topic. I added a PS symposium (Leech et
al. 2002) on elite interviewing with both overseas and domestic
examples. I also added a chapter (Walsh 2009) from the forth-
coming volume Political Ethnography edited by Edward
Schatz. Like Schaffer’s, Walsh’s piece offers students a tan-
gible connection to the in-person field research experience, in
her case by excerpting conversations of ordinary citizens of
Michigan. Finally I added selections from scholarly interviews
with Robert Bates and James Scott (Munck and Snyder 2007)
in which each talks about his field research overseas. My goal
here was both to show leading political scientists of very dif-
ferent methodological persuasions arguing that field research
is essential, and to give more concrete examples of the back-
and-forth dynamics of good interviewing.
Conclusion: Two Take-Home Lessons
Early in this contribution I suggested that instructors seek-
ing to pragmatically make room for interpretivism be self-criti-
cal about where they themselves are coming from, in order to
be as open as possible to learning from student reactions,
even, and indeed especially, reactions that cut against prior
beliefs. My first concluding lesson is that it is no less impor-
tant to reflect on where our students are coming from. Looking
back over the reactions reported above, I am constantly re-
minded that the bulk of my students were third-years, in the
middle of taking comprehensive exams, and anxious about for-
mulating a dissertation project that could engage faculty advi-
sors and, hopefully, in the longer term, political scientists else-
where. Having invested much time and effort in prior classes
and exam studying, students at this stage in a PhD program are
especially receptive to methods readings that connect to works
and debates they are already familiar with. This both makes a
reading more accessible and reassures students that the method
or perspective being presented is within the bounds of the
“political science” they are being socialized into. For any
scholar, faculty or student, our sense of what “political sci-
ence” is has been shaped by what we have read (or at least
read about), and third-year students have already read a lot!
Students at this stage are, moreover, also understandingly ea-
ger consumers of readings that give a concrete sense of, and
advice about, the practical realities of doing research.
My second concluding lesson concerns the results of
exposing students to novel perspectives or methods. Parti-
sans of interpretivism should be aware that increasing student
exposure is no sure route to greater disciplinary recognition,
let alone use, of interpretive perspectives and methods. Stu-
dents are busy people who allocate attention selectively. They
may skim the surface and not really engage interpretive read-
ings (especially if they are assigned plenty of other readings).
Alternatively, they may engage interpretivism, but do so via
Qualitative & Multi-Method Research, Spring 2009
criticism and identifying against it. My take-home point is cer-
tainly not that making room for interpretivism will always be
futile or outright counterproductive. But I want to advocate
the relative payoff of selectively focusing on interpretive read-
ings that address activities our students already expect to pur-
sue. They all plan to construct explanations, and many to con-
duct interviews of one sort of another. Interpretivists have
distinct viewpoints to offer regarding the standards for good
explanation in social sciences, and practical guidance to give
about methods—such as in-depth interviewing—that help meet
those standards. Interpretivism is, I would suggest, most likely
to win a receptive hearing among political science graduate
students, and the discipline more broadly, when it engages
matters of common and practical concern.
Notes
1
The political theory subfield is predominantly interpretive. Hence,
more precisely, the point in question is whether interpretivism should
be given more room in the methods training of graduate students
outside of political theory.
2
The flow from assumptions to arguments here is not automatic.
Advocates of greater exposure subscribing to a conversion assump-
tion also believe (or hope) that the consequences of conversion will
be positive for individual converts and, over the longer term, for the
discipline more broadly. Different assessment of consequences could,
however, reframe “conversion” as a “corruption,” either of individual
careers or the discipline.
3
At the end of the semester I gave students fresh copies of the
syllabus and had them put a + (or multiple ++’s for stronger reac-
tions) next to readings they would recommend for the next iteration
of PSc 209, and a – next to those they would not. I did not single out
interpretive readings for special attention in this process, but rather
encouraged students to rate any readings that stood out to them.
4
Student reactions informed changes to various aspects of my
syllabus, not just my interpretivism readings. Thus, for example, I
also dropped a unit on lab experiments in political science.
5
The tensions between “humanist” (aka “interpretive”) and “criti-
cal” philosophies deserve, I believe, just as much attention as the
disputes either of them has with “positivism.” Eliding these tensions
helps construct the appearance of a united “anti-positivist” position,
but that appearance is purchased at the price of philosophical confu-
sion.
References
Abbott, Andrew. 1998. “The Causal Devolution.” Sociological Meth-
ods and Research 27:2, 148–181.
Abbott, Andrew. 2004. Methods of Discovery: Heuristics for the
Social Sciences. New York: W.W. Norton.
Adcock, Robert and David Collier. 2001. “Measurement Validity: A
Shared Standard for Qualitative and Quantitative Research.” Ameri-
can Political Science Review 95:3, 529–546
Adcock, Robert. 2003. “What Might it Mean to be an ‘Interpretivist’?”
Qualitative Methods: Newsletter of the Qualitative Methods Sec-
tion 1:2, 16–18.
Adcock, Robert. 2009. “Making Making Social Science Matter Mat-
ter to Us.” Journal of Theoretical Politics 21:1, 97–112. .
Brady, Henry and David Collier, eds. 2004. Rethinking Social In-
quiry: Diverse Tools, Shared Standards. Lanham, MD: Rowman
& Littlefield.
Collier, David and Steven Levitsky. 1997. “Democracy with Adjec-
7
Qualitative & Multi-Method Research, Spring 2009
tives: Conceptual Innovation in Comparative Research.” World
Politics 49, 430–451.
Dryzek, John S. 1986. “The Progress of Political Science.” Journal of
Politics 48:2, 301–320.
Elster, Jon. 2007. Explaining Social Behavior: More Nuts and Bolts
for the Social Sciences. New York: Cambridge University Press.
Fay, Brian and J. Donald Moon. 1977. “What Would an Adequate
Philosophy of Social Science Look Like?” Philosophy of the Social
Sciences 7, 209–227.
Fenno, Jr., Richard F. 1977. “U.S. House Members in Their Con-
stituencies: An Exploration.” American Political Science Review
71:3, 883–917.
Fenno, Jr., Richard F. 1990. “The Political Scientist as Participant
Observer.” In Richard F. Fenno, Jr., Watching Politicians: Essays
on Participant Observation. Berkeley: IGS Press.
Geertz, Clifford. 1973. “Thick Description: Toward an Interpretive
Theory of Culture.” In Clifford Geertz, The Interpretation of Cul-
tures. New York: Basic Books.
Geertz, Clifford. 1979. “‘From the Native’s Point of View’: On the
Nature of Anthropological Understanding.” In Paul Rabinow and
William M. Sullivan, eds., Interpretive Social Science: A Reader.
Berkeley: University of California Press.
Godfrey-Smith, Peter. 2003. Theory and Reality: An Introduction to
the Philosophy of Science. Chicago: University of Chicago Press.
Goldthorpe, John H. 2001. “Causation, Statistics, and Sociology.”
European Sociological Review 17:1, 1–20.
Kurzman, Charles. 2004. “Can Understanding Undermine Explana-
tion? The Confused Experience of Revolution.” Philosophy of the
Social Sciences 34:3, 328–351.
Leech, Beth L., et al. 2002. “Interview Methods in Political Science.”
PS: Political Science and Politics 35:4, 663–688.
Mahoney, James and Gary Goertz. 2006. “A Tale of Two Cultures:
Contrasting Quantitative and Qualitative Research.” Political Analy-
sis 14, 227–249.
Munck, Gerardo L. and Richard Snyder. 2007. Passion, Craft, and
Method in Comparative Politics. Baltimore: Johns Hopkins Uni-
versity Press.
Neuman, W. Lawrence. 2006. Social Research Methods: Qualitative
and Quantitative Approaches, 6th ed. Boston: Allyn and Bacon.
Oren, Ido. 2006. “Political Science as History.” In Dvora Yanow and
Peregrine Schwartz-Shea, eds., Interpretation and Method: Em-
pirical Research Methods and the Interpretive Turn. Armonk, NY:
ME Sharpe.
Osborne, Thomas and Nikolas Rose. 1999. “Do the Social Sciences
Create Phenomena? The Example of Public Opinion Research.”
British Journal of Sociology 50:3, 367–396.
Roth, Paul A. 2004. “Hearts of Darkness: ‘Perpetuator History’ and
Why There is No Why.” History of the Human Sciences 17:2/3,
211–251.
Schaffer, Frederic Charles. 2006. “Ordinary Language Interviewing.”
In Yanow and Schwartz-Shea, eds., Interpretation and Method.
Schwartz, Joel D. 1984. “Participation and Multisubjective Under-
standing: An Interpretivist Approach to the Study of Political Par-
ticipation.” Journal of Politics 46:4, 1117–1141.
Schwartz-Shea, Peregrine. 2003. “Is this the Curriculum We Want?
Doctoral Requirements and Offerings in Methods and Methodol-
ogy.” PS: Poitical Science and Politics 36:3, 379–386.
Shehata, Samer. 2006. “Ethnography, Identity, and the Production of
Knowledge.” In Dvora Yanow and Peregrine Schwartz-Shea, eds.,
Interpretation and Method: Empirical Research Methods and the
Interpretive Turn. Armonk, NY: ME Sharpe.
Soss, Joe. 2006. “Talking Our Way to Meaningful Explanations.” In
Yanow and Schwartz-Shea, eds. Interpretation and Method. Em-
8
pirical Research Methods and the Interpretive Turn. Armonk, NY:
ME Sharpe.
Taylor, Charles. 1971. “Interpretation and the Sciences of Man.”
Review of Metaphysics 25, 3–51.
Yanow, Dvora and Peregrine Schwartz-Shea. 2002. “‘Reading’ ‘Meth-
ods’ ‘Texts’: How Research Methods Texts Construct Political
Science.” Political Research Quarterly 55:2, 457–486.
Yanow, Dvora and Peregrine Schwartz-Shea. 2006. “Introduction.”
In Yanow and Schwartz-Shea, eds. Interpretation and Method.
Empirical Research Methods and the Interpretive Turn. Armonk,
NY: ME Sharpe.
Walsh, Katherine Cramer. 2009. “Scholars as Citizens: Studying Pub-
lic Opinion through Ethnography.” In Edward Schatz, Political
Ethnography: What Immersion Contributes to the Study of Power.
Chicago: University of Chicago Press.
Wendt, Alexander. 1998. “On Constitution and Causation in Interna-
tional Relations.” Review of International Studies 24:5, 101–118.
Undoing the Opposition Between
Theory and Methods
Emily Hauptmann
Western Michigan University
emily.hauptmann@wmich.edu
The division of curricular labor in most departments usu-
ally leaves teaching methods courses to only a small portion
of the faculty. And in most departments, there is little overlap
between those who teach methods and those who teach politi-
cal theory. When I began teaching almost twenty years ago, I
would have ranked “methods” towards the bottom of courses
I expected or wanted to teach. But a few years ago, that began
to change. Since 2006, I have team-taught a graduate course in
Qualitative and Interpretive Methods and an undergraduate
course in Scope and Methods of Political Science. This spring,
I will offer The Logic of Political Inquiry, a graduate course on
the history of the discipline and the philosophy of the social
sciences. Several circumstances (which I touch on below) made
it possible for me to teach this array of courses.
I do not expect most theorists have either the inclination
or the opportunity to explore teaching any methods courses.
But I believe that this state of affairs is a product of the over-
drawn opposition between theory and methods that many of
us implicitly accept. In what follows, I discuss the conflicting
ways in which I learned about methods and what I now believe
are the best reasons for theorists to consider contributing to
their department’s methods offerings.
Learning “Theory versus Methods”
I learned a particular view of the discipline as a political
theory student at Berkeley in the 1980s. At the time, Berkeley’s
graduate curriculum included no general methods requirements;
like many other students in political theory, I finished my gradu-
ate education without any course work in either research meth-
ods or statistics. What I learned instead was to see what politi-
cal theorists did and knew in opposition to any approach that
stressed methods—what Sheldon Wolin famously called
 Qualitative & Multi-Method Research, Spring 2009
“methodism.” In his 1969 polemic against behavioralist politi- Learning Methods Without “Methods”
cal science, Wolin argued that methods-driven political sci-
At the same time I was learning that theorists were not
ence was inherently incapable of a critical perspective on po-
methodists, I learned methods all the same—ordinary lan-
litical life. The “methodist,” Wolin contended, necessarily privi-
guage analysis and conceptual analysis. (That neither I nor
leged regularity and order, her/his claims to objectivity not-
my teachers would have called them “methods” at the time
withstanding (1064). For a beginning theorist, this was an easy
shows how much disciplinary standards have changed.) Need-
argument to accept: it bolstered my sense of the grandeur of
less to say, there was no course in the curriculum called “Meth-
the work I wanted to do to imagine it in opposition to the
ods for Theorists”; nor did any particular class I took focus
willfully blinkered existence of Wolin’s methodist.
primarily on either ordinary language or conceptual analysis.
Some twenty years later, I now see several things about
Instead, I began to learn about these methods principally
Wolin’s essay that I missed when I first encountered it. For
through a few readings for Hanna Pitkin’s graduate seminars
one, as I have argued elsewhere, I now read the essay as a
and by studying her own work on representation and justice.
more specific defense of Wolin’s view of political theory against
I became more deeply immersed in ordinary language and con-
what the partisans of the behavioral revolution took theory to
ceptual analysis while working on my dissertation on rational
be (Hauptmann 2005). That debate about the nature and status
choice theory (which Pitkin directed). Ultimately, I decided to
of theory in political science continues—although, as I sug-
devote a chapter of my dissertation to an ordinary language
gest below, I believe political theorists could take part in it in a
analysis of “choice,” though I did so knowing I had a lot to
fuller way than many choose to do.
learn before I could do this kind of analysis well. My learning
Further, though Wolin pitches his attack against method-
happened slowly, through trial and error; I would try my hand
ism in general, I now read his critique as best directed against
at a bit of analysis to which Pitkin would respond and correct.
the positivist practices that came to dominate political science
I’d revise, she’d respond, and I’d revise again. The way I’ve
in the 50s and 60s. Those practices, Wolin compellingly ar-
summarized my learning experience above reveals one peda-
gues, allowed many in the discipline to present their work as
gogical challenge all of us who teach qualitative-interpretive
ideologically neutral, rigorous, and systematic—but at the cost
methods have to face. If we want to try teaching these meth-
of offering a critical perspective on political life. It seems ille-
ods in a class rather than one-on-one, how do we do so well—
gitimate, however, to find all work that relies on “methods”
without forcing them into a rigid sequence of “how to” steps?
guilty of this shortcoming. Indeed, far from assuming a world
(Yanow 2006: 70–72).
governed by “regularities” (as Wolin charges methodist politi-
cal scientists do [1064]), some of the methods we now employ Theorists Who Teach Methods? Finding the
are particularly well attuned both to identifying and making Right Space in the Curriculum
sense of irregularities (e.g., Oren 2003; Schaffer 1998, Yanow
Methods matter to me now not only as a researcher but
2003). Indeed, many who employ methods like category and
also as a teacher, advisor, and member of a department. That
ordinary language analysis or reflexive historical analysis would
is, I now think about how methods should figure in my
argue that these fuel rather than sap the critical power of their
department’s undergraduate and graduate curricula and what
work.
methods would best suit the projects my graduate students
Still, some elements of Wolin’s critique remain valid. In
pursue. Thinking about methods in these contexts inevitably
particular, Wolin noticed how the methods portion of curricula
means confronting how differently many of my colleagues
demanded huge investments of time on the part of both stu-
think about them. Although, for the reasons I touched on
dents and faculty—and warned that this would have conse-
above, it was easy to scorn thinking about methods while I
quences for everything else a department did (1064, 1072–73).
was a graduate student, as a faculty member I came to see that
This shift was just beginning when Wolin noticed it; most of
doing so barred me from pursuing the important pedagogical
us now see curricula that place a heavy emphasis on methods
path along which departments teach their students how to
as the norm. Of course, not all programs require all students to
think about politics. I do not mean to suggest that graduate
take a methods sequence. But the vast majority do (Schwartz-
seminars in political theory do not do this kind of teaching;
Shea 2005: 389). And to the extent political theorists do not
many do. But if methods courses are required of all students
participate in designing and teaching these courses, they end
and none of them are taught by political theorists, depart-
up ceding an important part of the work of defining the disci-
ments end up reproducing what I now see as the false oppo-
pline to their colleagues in other sub-fields. This isn’t always
sition between theory and methods.
the most immediately rewarding work; nor will every group of
So what is a theorist to do? In what follows, I want to
faculty primarily responsible for methods be happy to allow
point to some spaces in undergraduate and graduate cur-
others to join in. But for the reasons I discuss in more detail
ricula that are in principal open to theorists who wish to con-
below, I think it’s time for those of us who teach political theory
tribute to how their departments teach students how to think
to think about how we might chip away at the opposition be-
about politics and do research. Of course, not all of these
tween theory and methods by making a few forays into teach-
spaces will be open to every theorist at every institution. But
ing methods ourselves.
at least a few should be.
There are several types of undergraduate courses that

9
Qualitative & Multi-Method Research, Spring 2009
offer theorists some of the pedagogical space I have been
describing. One such course is commonly called either “Scope
and Methods of Political Science” or “Political Science as a
Discipline.” To the extent that such courses are not under-
stood primarily as “how to” courses in methods, those who
take them on might focus on the history of the discipline or on
ontological and epistemological issues raised by philosophers
of social science. Many theorists are especially well versed in
the latter approach.
Additionally, a significant number of departments also
require their graduate students to take at least one course in
the philosophy of science (Schwartz-Shea 2005: 388–389). Such
courses are not too far afield from methods courses and can
therefore speak to how to pose research questions as well as
the merits of various methods. Though all political theorists
do not have a strong grounding in the philosophy of science,
many find the kinds of epistemological questions raised by
philosophers of science both familiar and engaging. What’s
more, it’s not a huge step to go from showing what follows
from a particular theoretical world view to analyzing the epis-
temological and ontological underpinnings of several differ-
ent methodologies.
Finally, many departments offer courses in Research Meth-
ods both at the undergraduate and graduate levels. Theorists
might be able to teach a portion of a team-taught general
research methods or qualitative-interpretive methods course
or build their own course under the latter title. Many people
with whom I’ve spoken at recent Methods Cafés who teach
general research methods courses have specifically asked me
for suggestions on how to include qualitative-interpretive
methods into their courses. In some cases, therefore, even
general research methods courses offer theorists an opportu-
nity to contribute to how their department teaches methods.
A Theorist Teaches Methods—My First Attempts
Although I just began teaching methods over the last
few years, I have now done so at both the undergraduate and
graduate levels. By the end of this academic year, I will have
taught three courses that are at least in part devoted to meth-
ods. I hope that by summarizing my approach to several of
these I can add some specificity to the general point that
theorists can potentially contribute to how their departments
teach methods.
I first taught methods at the graduate level in a course
designed by myself and two of my colleagues (Susan
Hoffmann, who studies American public policy, and Sybil
Rhodes, a Latin Americanist) two years ago. We called the
course “Qualitative Methods,” but taught it under a special
topics number in our department’s graduate methods se-
quence. (For those who teach in programs that do not yet
have an established qualitative methods course, teaching one
under a special topics number is a good way to start.) We
divided the course into three principal sections:
(1) Debates about research design, inference and inter-
pretation;
(2) Issues surrounding conceptualization; and
(3) Approaches to data collection and analysis.
10
Since our backgrounds and current research differed substan-
tially, it was fairly obvious who ought to teach what. I led
several sessions on ordinary language and symbolic analysis
and presented my own work in a week devoted to working
with archival collections. Working with several colleagues in
this way to put a new course together was the only way we
could have begun, since none of us felt prepared to offer the
entire course on our own.
This semester I began teaching Scope and Methods of
Political Science at the undergraduate level. My coming to
teach this course and its place in my department’s curriculum
both require some explanation. The course became my respon-
sibility following the retirement of my colleague, Alan Isaak,
who joined the department in the mid-1960s. Hired primarily to
teach courses in the history of political thought, Alan was
also asked by the department to develop a Scope and Meth-
ods course. He first taught the course in the late 1960s and
soon after published the first edition of Scope and Methods of
Political Science, a widely used textbook now in its fourth
edition. Alan studied both empirical political theory and the
philosophy of science as a graduate student at Indiana; both
subjects were central to the Scope and Methods course he
taught for nearly 40 years.
My taking on this course after Alan’s retirement made
sense, given its history in my department. Teaching it, how-
ever, has put me into a position I have never been in before:
for the first time, I am teaching a course specifically required
of all political science majors. (Although my department re-
quires its majors to take one course in the history of political
thought, none of the four such courses we offer is specifically
required.) Reflecting on this unprecedented pedagogical op-
portunity has led me to see both how central methods courses
can be to defining students’ sense of what political science is
and how peripheral many political theory courses are to doing
so. I want to be clear: I am not saying political theory courses
cannot speak to these issues, but rather that their place in
most undergraduate curricula does not make this one of their
primary purposes. Speaking to students’ sense of what politi-
cal science is, however, is what Scope and Methods is all
about.
As with any course one is teaching for the first time, I
doubt I have hit upon the best way of teaching Scope and
Methods on my first try. I have tried to build the course around
issues I think are important for all majors to encounter without
making it into too sweeping a survey. To that end, I constructed
the course around several focal points: U.S. political science
in history; sciences of politics; and, for the election season,
disagreements about how to study voting and elections. I am
closing the course by devoting three weeks to research pre-
sentations, one by myself and two by my colleagues. After
students have read portions of our recent research, each of us
will discuss the methods we have used in it.
The most challenging aspect of teaching this course has
been finding the best way to encourage students to think
theoretically. I believe it’s not so much the case that they
would find thinking this way difficult in all contexts; rather,
they are puzzled by being asked to think this way in a politi-

cal science class. I wish I could say that students who have
already taken a political theory course seem better prepared
for the class; but that does not seem to be the case. In my
experience, many students regard courses in the history of
political thought as curious addenda to the political science
curriculum that lead few to think about the discipline differ-
ently. By contrast, I’ve had an easier time convincing my stu-
dents in Scope and Methods that what they are learning is
connected to what they have learned in other courses. But
taking the next step—trying to provide my students with a
conceptual vocabulary to use to think about how they have
learned political science—has proven the most difficult. For
instance, although students can readily grasp “positivism”
as a general conception of science, it’s considerably more
difficult for most to explain why a particular piece is or is not
positivist—or to say what sorts of things someone working
within a positivist tradition is likely to overlook or dismiss.
Though I cannot say I have hit upon a good way to do this
yet, I see teaching this class as the best opportunity I have
yet had to weave theoretical thinking into how students un-
derstand political science.
This coming spring, I will be teaching the Logic of Politi-
cal Inquiry, a graduate-level course developed and taught for
many years my colleague, Alan Isaak. Because I believe that
current standards for legitimate political inquiry are the result
of recent disciplinary history, I plan on structuring the course
around a number of intradisciplinary debates over what counts
as sound political science. I intend to tie our discussions of
the philosophy of the social sciences explicitly to that struc-
ture as well.
Teaching such an array of methods courses is unusual
for a political theorist. I have tried to speak from my experi-
ences not only to undo the opposition between theory and
methods but also to show how its reproduction excludes po-
litical theorists from participating in part of what political sci-
ence faculties do to teach their students to think about poli-
tics.
References
Hauptmann, Emily. 2005. “Defining ‘Theory’ in Postwar Political
Science.” In The Politics of Method in the Human Sciences: Positiv-
ism and its Epistemological Others. George Steinmetz, ed. Chapel
Hill, NC: Duke University Press.
Isaak, Alan C. 1997. Scope and Methods of Political Science: An
Introduction to the Methodology of Political Inquiry, 4th edition.
Orlando, FL: Harcourt Brace & Company.
Oren, Ido. 2003. Our Enemies and US: America’s Rivalries and the
Making of Political Science. Ithaca, NY: Cornell University Press.
Schaffer, Frederic Charles. 1998. Democracy in Translation: Under-
standing Politics in an Unfamiliar Culture. Ithaca, NY: Cornell
University Press.
Schwartz-Shea, Peregrine. 2005. “The Graduate Student Experience:
‘Hegemony’ or Balance in Methodological Training?” In Perest-
roika! The Raucous Rebellion in Political Science. Kristen Renwick
Monroe, ed. New Haven: Yale University Press.
Wolin, Sheldon. 1969. “Political Theory as a Vocation.” The Ameri-
can Political Science Review 63, 1062–1082.
Yanow, Dvora. 2003. Constructing American “Race” and “Ethnicity”:
Qualitative & Multi-Method Research, Spring 2009
Category-Making in Public Policy and Administration. Armonk,
NY: M.E. Sharpe.
Yanow, Dvora. 2006. “Neither Rigorous Nor Objective? Interrogat-
ing Criteria for Knowledge Claims in Interpretive Science.” In In-
terpretation and Method: Empirical Research Methods and the
Interpretive Turn. Dvora Yanow and Peregrine Schwartz-Shea, eds.
Armonk, NY: M.E. Sharpe.
Introducing Rigor to the Teaching of
Interpretive Methods
Jan Kubik
Rutgers University
kubik@rci.rutgers.edu
I offer several remarks on the way I introduce interpretive
methods within a course on research design and intend to
approach a new course devoted to a more systematic exposi-
tion of interpretivism and its methods. Due to space limita-
tions, I signal the key issues that must be addressed and offer
two examples of specific analyses, as the devil tends to be in
details.
Interpretation, an intellectual operation whose essence
has been variously rendered as translation, clarification, or
placing in context, can be usefully introduced to students
with the question, “What needs to be interpreted during the
course of a social scientific inquiry?” In part, the answer to
this question involves recognizing when we are interpreting.
How are the concepts in our studies operationalized? Are
their meanings transparent and universal or do they vary
across cultural contexts? What accounts for the processes
evidenced in formal modeling? To some degree all social sci-
ence analysis, whether aimed at classifying a phenomenon,
imputing a cause, or articulating a process involves interpre-
tation. But in another important sense interpretation is a spe-
cific method of understanding the communicative process
through which discursive objects are created, contested, em-
ployed, and interpreted (by actors) as part of the machinery of
power.
Why Interpret?
While the utility of interpretivist approaches is taken for
granted in anthropology, sociology, cultural studies, or femi-
nism, it is far from obvious to many practitioners of political
science. The reasons for this may be complex, but they seem
to be rooted in the predominantly naturalistic tenor of the
discipline (for an alternative perspective see Chabal and Daloz
2006; Smith 2004).1 But if we agree with an (anti-naturalist)
assumption that the signifying process through which people
build models of the world, particularly of the social and politi-
cal world, has political relevance, then the study of how such
models are constructed, transmitted, maintained, and received
becomes of interest to us. The study of such issues is incon-
ceivable without interpretation. Among the phenomena rou-
tinely studied with the help of interpretive approaches are, for
example: (1) legitimacy (as its standards vary from society to
society and depend on contextualized, culture-specific crite-
11
Qualitative & Multi-Method Research, Spring 2009
ria [Aronoff 1991, Kubik 1994]); (2) mechanisms of compli-
ance, quiescence, and everyday resistance (as they involve
the manipulation of cultural understandings of reality
[Wedeen 1999]); and (3) generation, reproduction, and dis-
mantling of collective identities (as such processes involve
the use of “cultural materials” [Ross 2007, Fernandes 2006,
Davis 2005]). Interpretive approaches also provide fresh and
valuable insights into a number of problem areas usually stud-
ied through naturalistic modes of inquiry. Ample evidence of
the fruitfulness of interpretation can be found in the study of
non-Western political systems in comparative politics (Chabal
and Daloz 2006, Ashforth 2005), constructivist work in inter-
national relations (Kratochvil and Ruggie 1986, Klotz and Lynch
2007), “culturalist” analyses in political economy (Blyth 2002),
and even economics (Rao and Walton, eds. 2004).
The first task in introducing interpretation to students,
then, is to get them to recognize that interpretation underlies
the entire enterprise of social science. Whether we are con-
ceptualizing a particular problematic, operationalizing con-
cepts, or organizing observations for analysis, there is a di-
mension of interpretation. I try to demonstrate the reliance of
fundamental concepts of political science analysis on a pro-
cess of signification that is not necessarily transportable from
one social context to another. Cognitive, symbolic, linguistic,
and communicative dimensions of political processes vary
and are clearly influenced by the historical experiences of the
particular society. Recognizing that one is always interpreting
is a critical aspect of acquiring methodological self-aware-
ness. The question then is twofold: how interpretation sup-
ports or relates to other methods and how it is employed in a
rigorous and systematic manner that meets the standards of
scientific analysis we presume are achieved through these
other methods.
Location of Interpretivism among Other Approaches
Interpretivists are beginning to systematize their approach
and engage in an explicit exposition of its assumptions, meth-
ods, and techniques (Yanow and Schwartz-Shea, eds. 2006).
Situating interpretivism vis-à-vis the predominant qualitative
and quantitative methodologies illuminates the basic onto-
logical and epistemological assumptions that underpin the
Table 1: Ontological and Epistemological Dualism
Unity of Object
(Naturalism)
Unity of Method Positivism
Epistemology
Different Methods ? Interpretivism
12
different approaches to knowledge production. Table 1 is a
simple tool I use to discuss the ontological split between
naturalists and anti-naturalists with my students. I begin by
introducing the Geisteswissenschaften versus Naturwissen-
schaften distinction of Droysen and Dilthey and then ask
them to reflect on the proposition that anti-naturalism is asso-
ciated with a specific ontology of the social and therefore
calls for at least partial reliance on a specific method: interpre-
tation or understanding. The debate on the “proper” match
between methods and problems (Bevir and Kedar 2008) is a
fruitful point of departure for considering both the informa-
tion necessary to answer certain questions and the poten-
tially complementary relationship between different modes of
inquiry. For example, interpretivism can enhance survey work
(Stoker 2003: 13–16) or game theory (O’Neil 1999, Bates et al.
1998, Johnson 2002).
Interpretation of What?
Three Basic Varieties of the Operation
There are at least three dimensions of political science
research that engage interpretive skills, whether consciously
or not. The first is in recognizing and classifying observations
(or “data”). For example, is a group of people gathered in a
market square a religious procession, a political rally, or a crowd
getting ready for an open-air concert? Does the uniform of a
person whose actions we are studying signify a soldier or a
miner? Interpretive skills enable basic coding and classifica-
tion. Without them, much comparative work is inconceivable.
Weber calls this type of interpretive work direct observational
understanding. I refer to it as classificatory interpretation.
The second interpretive moment comes when we try to specify
what drives human agency: “Why does/did she do this?” When
researchers ascribe motives (psychological approaches) or
reasons (rational choice approaches) to human behavior, they
engage in what Weber refers to as explanatory understanding.
I call it motivational interpretation. The third is in reconstruct-
ing the meaning of actions, statements, displays, performances,
etc. Discerning “What does she mean by this?” or “What is
the meaning of this action?” involves semiotic/communica-
tive interpretation.
Ontology
Different Object
(Anti-Naturalism)
Ontological
Dualism/Epistemological
Monism

Combining Motivational and Semiotic Interpretation
and Causal Explanation
Systematic methodological reflection about the relation-
ship between causal explanation and interpretation (under-
standing) began with Weber. My course thus takes Weber’s
methodology, including his own writings and several critical
commentaries on his seminal studies, as the point of depar-
ture. In addition, I present a sequence of slides designed to
elucidate the difference between motivational and semiotic
modes of interpretation, and to locate the latter within the
explanatory sequence. I rely on an example offered by Martin
(2000).
Simple Causal Relationship
A --------------------- B
If A, then B.
A – “Being the inhabitant of a city that is devastated by bombs”
B – “The lack of resistance to aggression”
Source: Martin (2000)
We begin with a simple model of a causal relationship.
The first step in analysis is to formulate a proposition that
captures the nature of the relationship between two social-
level phenomena. Such a proposition can be expressed thus:
“If a city is bombed, its inhabitants will not resist further
aggression.”2 Explanation generally involves more than a
simple statement of cause and effect, however. The link be-
tween them must be articulated by demonstrating why or how
a presumed cause produces a presumed effect. In an influen-
tial admonition to specify the underlying mechanisms in causal
explanations—effectively linking “macro” to “micro” expla-
nations—Coleman (1990) proposed what has come to be
known as the “Coleman boat” (Coleman-Lindberg diagram).
This model, illustrated below, inserts between two “macro”
phenomena and at least one “micro” mechanism.
From Macro to Micro: Coleman’s “Boat”
A --------------------- B helps to analyze the collective action dilemma and investigate
A1 B1
? “empathy,” but usually we employ a deductively constructed
A – macro causes A1 – micro causes
B – macro causes B1 – micro causes
Qualitative & Multi-Method Research, Spring 2009
Micro causes and micro effects are states or attributes of
individual people. How do we know whether A1 “causes” B1?
How do we get into “their heads”? What method or research
procedure do we need to discern why people act as they do?
One answer may come from empathy, understood roughly as a
combination of introspection and reasoning through analogy
(what transpires in my psyche can be attributed to other hu-
man beings).3 The next image demonstrates location of empa-
thy in the construction of motivational explanations.
(1) Weber’s Explanatory Verstehen
(According to Martin 2000: 18–25)
A --------------------- B
A1 B1
Empathy ? Verstehen?
A1 – “Feeling of terror and dread”
B1 – “Feeling of helplessness”
Empathy generates a bridging proposition that states, for
example: “If an individual feels terror and dread, as a result
(s)he also develops a feeling of helplessness.” Then, a testable
set of propositions can take us from the “experience of being
bombed” to the “feeling of terror and dread” connected through
empathy to the “feeling of helplessness” experienced individu-
ally, and finally to the “lack of resistance” by the whole/major-
ity of/a part of the population. The bridge between A1 and B1
can be built in many ways, without empathy, but with the use
of psychological theories about motivation. I do not develop
this here for lack of space.
We can also model the argument in game theoretic terms.
For example, statements about “feelings” can be replaced with
statements about “strategic calculations,” including a recon-
struction of an individual preference ordering and assessing
the viability of various courses of action (given the assumed or
observed actions of others).
A well-known benefit of game theoretic analysis is that it
under what conditions rationally calculating individuals en-
gage or do not engage in (collective) action given what they
know about the actions of others. A researcher may identify
“tipping points,” “cascading effects,” etc. Interpretation is
nonetheless a critical component of linking individual agency
to observed collective behavior by attributing “reasons” (rather
than “motivations) to individuals. In this task we may rely on
model of a “calculating, rational individual.” One way or an-
other, however, we begin our analysis by interpreting the mo-
tives or reasons “causing” individual (in)action.
13
Qualitative & Multi-Method Research, Spring 2009
(2) Rational Calculation
A --------------------- B effective and why?), from what position (was this person pow-
Collective action
dilemma: cascad-
A1 B1 ing effect and
tipping points
Reconstruction of strategic decision making
A1 – Evaluating various preferences for courses of
action (represented by a utility function)
B1 – Responding strategically (rationally) to actions
of others
Now, let’s imagine a modified situation. After a bombing
that most likely contributes to the lowering of the “fighting
spirit” of the population, a popular leader goes public (in all/
some available media) with a story that recounts the city’s
heroic past. She or he reminds people that once before their
ancestors successfully mobilized after an initial defeat, asks
them to overcome their fears and despondency, and appeals to
them for a common action against the enemy. In short, the
leader offers a narrative of empowerment. People listen; some
of them redefine their situation and begin to see it in a more
positive light. They convince themselves that success is pos-
sible or that they want to defend their city even if the chances
of success are miniscule. Mobilization for defense can be quite
effective among this section of the population.
(3) Intepretive Turn
A ----------------------------- ~B
A1 N ~B1
Intepretation
A1 – “Feeling of terror and dread” ~B1 – Feeling of power
N – Narrative of empowerment ~B – Resistance to
aggression
Once a narrative of empowerment begins to circulate
among the members of the group, it influences their con-
ceptualizations of the situation and as a result it may change
their motivations (or calculations). The story is a crucial ele-
ment of the analyzed situation. Many researchers will con-
clude that an account of what happened that does not include
an interpretation of the narrative of empowerment and an analy-
sis of its effectiveness is incomplete. They will want to know
the details of the story, its origins, and how it fits within the
14
broader context of the group’s culture. Often they will want to
know more: who told the story, through what channel(s) of
communication (and if there were several, which was most
erful?), to whom, and with what effect (were there groups in
the population who responded more readily to the appeal than
others and why?)? As the next image suggests, interpretively
oriented researchers want to study the semiotic practices that
shape individual motivation and rational calculation rather than
the message alone (Wedeen 2004).
(4) Intepretation “In Context” (of Power and Institutions)
A ----------------------------- ~B
A1 N ~B1
Analysis of
semiotic practices
Source
The insertion of analyses of semiotic practices into studies of
international conflicts, electoral campaigns, or strategies of
political resistance is increasingly common; it is particularly
intriguing in the work of developmental economists (Rao and
Walton, eds. 2004).
Interpretation and the Study of Semiotic Practices
The study of semiotic practices can be designed in many
ways. For example, it can be fashioned as a reconstruction of
an event in terms of a Turnerian4 social drama (Wagner-Pacifici
1986), an ethnographic case study (Geertz 1973), ethnographi-
cally grounded sociology (Wedeen 1999), policy analysis
(Yanow 1997), or game theoretic modeling (O’Neill 1999). It can
be grounded in the vocabulary and models proposed by po-
litical economists or “pure” institutionalists. All of these ap-
proaches, however, involve an interpretive component.
There are many ways to begin studying semiotic/com-
municative interpretation and its components. Umberto Eco
breaks down the process into three tasks. First, we may want
to identify the intended meaning of the message (text, dis-
course, poster, painting, speech, performance, etc.). Intentio
auctoris—as Eco calls it—is not always available and usu-
ally difficult to reconstruct. Nonetheless, a skillful interviewer
or diligent biographer may shed a light on this component of
cultural creation. Second, we need to analyze the meaning(s)
of the message, intentio operis. This is the proper subject of
semiotic analysis and structural work. Methods of reconstruct-
ing syntagmatic chains (how to build “sentences” of culture?),
paradigmatic sets (what are building blocs of cultural forms
appropriate for a specific task?), and pragmatic strategies (what
is more likely to “work” in a given population?) are described
in countless manuals, including works on content and dis-
course analysis (for a useful introduction of some key issues,
 Qualitative & Multi-Method Research, Spring 2009
see Herrera and Braumoeller 2004). Third, the study of recep- the title “father of the fatherland” could only be applied
tion, the interpretation of the message by the (intended or to an archpastor-bishop and primarily to the patriarch…
unintended) audience, is a critical aspect of understanding therefore the designation in question could be inter-
public communication. Reception can vary substantially within preted as meaning that Peter was head of the church and
a given group and in many cases reflects an active process of proclaimed himself a patriarch. And that is precisely how
resistance through deliberate re-characterization of the mes- it was interpreted. (Uspenskij 1977:109)
sage. Intentio lectoris can be studied through in-depth inter-
First, by using the word “assumed,” Uspienskij signals that
viewing, participant observation, and surveys. This last tool is
he practices a post-structuralist style of analysis. It calls for
routinely used in the approach called political culture. It is
the identification of agency and its actions and is founded on
important to remember that the study of attitudes, orienta-
an assumption that cultural change is not a matter of apersonal
tions, and perceptions contributes to the reconstruction of
transformations, as it was usually modeled in structural analy-
only one dimension of the cultural process. A complete analy-
ses, but, rather, is caused by deliberate actions of specific
sis of this process must include two other elements: the recon-
actors. Second, the word “translation” identifies a specific
struction of the authors’ intentions (particularly intended mean-
semiotic operation. Uspienskij identifies the source of the
ings) and the study of the message itself.
“translated” cultural idea: the classical Rome. Third, the con-
If we agree that the study of politics should encompass
text (“Russian”) into which the translated element is inserted
the analysis of the communicative (cultural) process through
is identified and its transformative capacity is emphasized.
which some actors propose certain world-views, encode them
Fourth, the mechanism of (potential) semiotic transformation
in symbolic vehicles, and try to disseminate them, while others
is actually (albeit briefly) described. The analyst, in this case
encode and interpret these messages and accept, reject, or
an “expert” on the Russian culture, identifies the field of po-
simply register their “meaning,” then the study of the “sym-
tential meanings of the new element and points to the mean-
bolic content” of the messages is unavoidable.5 Only if we
ing that is privileged by the logic of culture. I believe that the
understand the message and its place in a broader cultural
interpretive (or semiotic) analysis is seriously flawed if such
context can we deduce its political significance. Each cultural
logic is not reconstructed independently from the reconstruc-
product is formed by its author, who selects from a rich albeit
tion of actors’ actual (interpretive or otherwise) choices and
not infinite repertoire of available cultural materials. If we want
actions. Fifth, Uspienskij informs the reader how the new cul-
to understand the strategy (politics) behind such choices we
tural idea (Peter as pater patriae) actually began to function:
need to be able to contrast choices that are actually made with
the popular interpretation followed the path privileged by the
the options that are (deliberately or not) forgone. That is why
cultural logic predominant at that time in Russia. He does not
interpretation needs to go beyond merely determining if a cul-
tell us how he knows this (his historical studies?), but we can
tural message does or does not have a causal effect (King,
easily imagine the utility of modern survey instruments in
Keohane and Verba 1994: 36–41). In most cases, we want to
arriving at such a conclusion.
know how a semiotic practice works and why it is (in)effective.
One of the untapped (by political scientists) reservoirs of Misconceptions and Myths: Interpretation Clarified
high-quality interpretive work is the Russian (and Soviet)
The perception of interpretive methods in political sci-
school of semiotics. Its authors have long recognized that
ence is fraught with misconceptions. Some are based on the
cultural mechanisms—together constituting a huge coordi-
lack of knowledge or erroneous understanding of what “inter-
nating system—need to be carefully studied in order to im-
preters” do. Some have deeper philosophical roots and are
prove our understanding of politics and, in particular, the
related to misplaced specifications of differences among vari-
machinery of power. A study of how Peter the Great built
ous epistemological positions (see Bevir and Kedar 2008 for
legitimacy for his power, by Boris Uspienskij, is exemplary of
some important comments on this issue). For lack of space, I
carefully crafted interpretive work. In the following passage
will only signal several problems belonging to the first group.
drawn from Uspienskij’s analysis, italicized and boldfaced
First, it is sometimes asserted that interpretive methods
words or phrases signal critical stages in the interpretive pro-
are non-empirical (Ragin 1987: 3, 35). The validity of this argu-
cess, which are briefly discussed below:
ment rests, of course, on the definition of empiricism. Without
In 1721, Peter assumed a new title: he began to be offi- entering a complex philosophical debate, it may be advisable
cially called “Emperor,” “the Great,” and in addition, “fa- to offer students some readings from art history and discuss
ther of the fatherland.”… This expression is nothing other with them the meaning of “empiricism” in concrete interpre-
than a translation of the Latin pater patriae, an honor- tive studies. I often recommend studying Ervin Panofsky’s
ary title of the Roman emperors. However, it had a differ- method of iconological analysis (1972) and it application in a
ent ring in a Russian cultural context. Since paternity in short work on Gothic architecture and scholasticism (1951).
general can be either blood or spiritual kinship, and Peter Second, interpretation is sometimes presented as an intel-
obviously could not be the people’s father in the sense lectual operation based only on empathy or introspection. It is
of blood kinship, this name was understood to be a pre- not difficult to show that this is simply not true. Dilthey, one of
tension to spiritual kinship. But only a member of the the founding figures in the history of hermeneutics, moved
church hierarchy could be a spiritual father, and in turn, ahead from “psychological” introspection to intersubjectively

15
Qualitative & Multi-Method Research, Spring 2009
verifiable reconstruction of cultural meanings. My favorite
method of dispelling this misconception is to have students
re-read Weber’s The Protestant Ethic and the Spirit of Capi-
talism. I ask them to find examples of “empathy.” None is to be
found. Weber works like a semiotician employing a form of
content analysis (admittedly rudimentary by today’s stan-
dards). He reconstructs and interprets the meanings of, say,
Franklin’s sayings. His interpretations are contestable and fal-
sifiable.
Third, critics claim that interpretation is imprecise, impres-
sionistic, undisciplined, and arbitrary. It is, however, easy to
demonstrate that in specific areas of interpretive scholarship
there exist specifiable rules, accumulation of knowledge, meth-
ods of achieving (and challenging) consensus, and inter-sub-
jective checks and balances. A useful way of introducing these
issues is to study the debate initiated by Geertz’s celebrated
interpretive essay on the Balinese cockfight, not merely the
essay itself (Jones 1998, Martin 1993, Segal 1999). Eco (1992)
offers another useful primer. In his exchange with Rorty on the
limits of interpretation, he warns against overinterpretation,
shows how we may try to avoid it, introduces a useful distinc-
tion between interpretation and use, and demonstrates that
the former is not completely arbitrary.
Fourth, semiotic interpretation is often seen as an induc-
tive mode of inquiry. Here, three arguments are worth fielding.
First, students may be asked to examine semiotic or hermeneu-
tic studies that focus on the way standardized and prescribed
methods of meaning encoding are realized in practice. An ex-
ample may be the study of the way various artists employ
prescriptions codified in manuals of allegorical and symbolic
representations (see Ripa 1971). Similar “manuals,” though
usually rather less explicit and precise, exist in other areas, for
example in the field of political advertising. The method em-
ployed in the examination of the “fit” between a manual’s in-
structions and specific realizations is at least partially deduc-
tive. Second, much interpretive work is founded on the falsifi-
cationist logic of conjectures and refutations a la Popper rather
than induction. Third, it can be argued that the logic of inter-
pretation is neither deductive nor inductive. It is abductive in
the Piercean sense. In turn, reflection on the logic of abduc-
tion, understudied by comparison with deduction and induc-
tion, helps to grasp the benefits and pitfalls of the critical
Popperian distinction between the context of discovery and
the context of justification, and inject some rigor into thinking
about the former.6
Fifth, given the difficulty of separating the context of jus-
tification from the context of discovery in interpretation (its
abductive character), another criticism, that interpretation is
good only for hypothesis generation and not for verification
or falsification, is misdirected.
Sixth, it is sometimes argued that interpretivists see inter-
pretation as the only goal of social science. Geertz’s famous
words that the analysis of culture is “not an experimental sci-
ence in search of law but an interpretive one in search of mean-
ing” (1973: 5) are often quoted in support of this view. Geertz’s
views evolved and grew more complex in his several decades
of work following that statement (see, for example, 1983; 2003:
16
27). But more importantly, many intepretivists do pursue causal
explanations and see interpretation as one research procedure
among many.
Seventh, interpretation is said to be unscientific. How-
ever, interpretation is arguably no less “scientific” than causal
inference. The relative status of either task depends on the
definition of science. For King, Keohane and Verba (1994),
“good research, that is, scientific research” (7) has four char-
acteristics:
(1) The goal is inference. There are two types of infer-
ence: descriptive and causal. Descriptive inference in-
volves “using observations from the world to learn about
unobserved facts.” Causal inference involves “learning
about causal effects from the data observed” (8)
(2) The procedures are public
(3) The conclusions are uncertain
(4) The content is method
Interpretation meets all four criteria: (1) it relies on inference to
connect observed phenomena (signifying elements) with the
(unobserved) meanings (signified elements); (2) its procedures
are (or at least are supposed to be) public and repeatable; (3)
its result are provisional (uncertain) and always subject to
verification and updating; and (4) its content can be con-
strued as method.
The task, whose realization has already begun, is to sys-
tematically demonstrate the validity of these points as well as
specify and examine the method’s:
(1) ontological affiliations (How are society and politics
understood and defined?);
(2) epistemological commitments (How are societies and
politics defined in a specific manner knowable?);
(3) rules and procedures;
(4) disciplinary varieties (semiotics, hermeneutics); and
(5) specific techniques (for example, content analysis,
[critical] discourse analysis, ethnographic accounts of
meaning-formation through rituals, etc.).
Notes
1
These remarks are mostly based on and related to the field of
comparative politics, my area of academic specialization. Amy Linch’s
assistance in sharpening my argument was invaluable. I also thank
my colleagues and students at Rutgers, who provided many critical
remarks while listening to my early efforts to develop this essay.
2
There are, of course, many ways to finesse this proposition,
express it in probabilistic terms, offer clearer conceptualizations of
the key concepts, etc.
3
It is easy to trivialize the role of empathy in Dilthey’s or Weber’s
methodologies. Martin (2000) provides a very useful discussion of
how empathy relates to other components of the understanding
(Verstehen) method.
4
Victor Turner’s approach and methods are clearly presented in
Turner (1974).
5
“The struggle over world views should itself be treated as a
strategic process” (Bates et al. 1998: 633–635).
6
For a useful, brief introduction to these issues, see Uve Wirth
(http://user.uni-frankfurt.de/~wirth/inferenc.htm), who observes:
“The Peircean account of abductive inference denies the possibil-

ity to draw a sharp borderline between ‘context of discovery’ and
‘context of justification.’”
References
Aronoff, Myron. 1991. Israeli Visions and Divisions: Cultural Change
and Political Conflict. New Brunswick, NJ: Transaction Publish-
ers.
Ashforth, Adam. 2005. Witchcraft, Violence, and Democracy in South
Africa. Chicago: University of Chicago Press.
Bates, Robert H., Rui J. P. de Figueirdo, Jr., and Barry R. Weingast.
1998. “The Politics of Interpretation: Rationality, Culture, and
Transition.” Politics and Society 26:4, 603–642.
Bevir, Mark and Asaf Kedar. 2008. “Concept Formation in Political
Science: An Anti-naturalist Critique of Qualitative Methodology.”
Perspectives on Politics 6:3, 503–517.
Blyth, Mark. 2002. Great Transformations: Economic Ideas and
Institutional Change in the Twentieth Century. Cambridge: Cam-
bridge University Press.
Chabal, Jean-Pascal and Patrick Daloz. 2006. Culture Troubles: Poli-
tics and the Interpretation of Meaning. Chicago: University of Chi-
cago Press.
Coleman, James. 1990. Foundations of Social Theory. Cambridge,
MA: Harvard University Press.
Davis, Eric. 2005. Memories of State: Politics, History, and Collec-
tive Identity in Modern Iraq. Berkeley: University of California
Press.
Eco, Umberto (with Richard Rorty, Jonathan Culler, and Christine
Brook-Rose). 1992. Interpretation and Overinterpretation. Stefan
Collini, ed. Cambridge: Cambridge University Press.
Fernandes, Leela. 2006. India’s New Middle Class: Democratic Poli-
tics in an Era of Economic Reform. Minneapolis: University of
Minnesota Press.
Geertz, Clifford. 1973. The Interpretation of Cultures. New York:
Basic Books.
Geertz, Clifford. 1983. Local Knowledge: Further Essays in Inter-
pretive Anthropology. New York. Basic Books.
Gerring, John. 2003. “Interview with Clifford Geertz.” Qualitative
Methods (Fall), 24–28.
Herrera, Yoshiko M. and Bear F. Braumoeller. 2004. “Symposium:
Discourse and Content Analysis.” Qualitative Methods (Spring),
15–39.
Johnson, James. 2002. “How Conceptual Problems Migrate: Ratio-
nal Choice, Interpretation, and Hazards of Pluralism.” Annual Re-
view of Political Science 5, 223–248.
Jones, Todd. 1998. “Interpretive Social Science and the ‘Native’s
Point of View’: A Closer Look.” Philosophy of the Social Sciences
28, 32–68.
King, Gary, Robert O. Keohane, and Sidney Verba. 1994. Designing
Social Inquiry: Scientific Inference in Qualitative Research. Prince-
ton: Princeton University Press.
Klotz, Audie and Cecelia Lynch. 2007. Strategies of Research in
Constructivist International Relations. Armonk, NY: M. E. Sharpe.
Kratochwil, Friedrich and John Gerard Ruggie. 1986. “International
Organization: A State of the Art on an Art of the State.” Interna-
tional Organization 40:4 (Autumn), 753–775.
Kubik, Jan. 1994. The Power of Symbols Against the Symbols of
Power: The Rise of Solidarity and the Fall of State Socialism in
Poland. University Park: Penn State University Press.
Martin, Michael. 1993. “Geertz and the Interpretive Approach in
Anthropology.” Synthese 97, 269–286.
Martin, Michael. 2000. Verstehen: The Uses of Understanding in
Social Science. New Brunswick, NJ: Transaction Publishers.
Qualitative & Multi-Method Research, Spring 2009
O’Neill, Barry. 1999. Honor, Symbols, War. Ann Arbor: University
of Michigan Press.
Panofsky, Ervin. 1972 (1939). Studies in Iconology: Humanistic
Themes in the Art of the Renaissance. New York: Harper and Row.
Panofsky, Ervin. 1951. Gothic Architecture and Scholasticism: An
Inquiry into the Analogy of the Arts, Philosophy, and Religion in the
Middle Ages. New York: Meridian Books.
Ragin, Charles. 1987. The Comparative Method: Moving Beyond
Qualitative and Quantitative Strategies. Berkeley: University of
California Press.
Rao, Vijayendra and Michael Walton, eds. 2004. Culture and Public
Action. Palo Alto, CA: Stanford Social Science (an imprint of
Stanford University Press).
Ripa, Cesare. 1971 (1758–60). Baroque and Rococo Pictorial Imag-
ery. New York: Dover Publications.
Ross, Marc Howard. 2007. Cultural Contestation in Ethnic Conflict.
Cambridge: Cambridge University Press.
Segal, Daniel. 1999. “A Response to Jones’s Critique of Interpretive
Social Science.” Philosophy of the Social Sciences 29, 306–309.
Smith, Rogers. 2004. “The Politics of Identities and the Tasks of
Political Science.” In Problems and Methods in the Study of Poli-
tics. Ian Shapiro, Rogers Smith, and Tarek E. Masoud, eds. Cam-
bridge: Cambridge University Press, 42–66.
Stoker, Laura. 2003. “Is it Possible to do Quantitative Survey Re-
search in an Interpretive Way?” Qualitative Methods 1:2, 13–16.
Turner, Victor. 1974. Dramas, Fields, and Metaphors: Symbolic Ac-
tion in Human Society. Ithaca: Cornell University Press.
Uspenskij, Boris A. 1977. “Historia sub specie semioticae.” In Soviet
Semiotics. Daniel P. Lucid, ed. Baltimore: Johns Hopkins Univer-
sity Press.
Wagner-Pacifici, Robin Erica. 1986. The Moro Morality Play: Ter-
rorism as Social Drama. Chicago: University of Chicago Press.
Wedeen, Lisa. 2004. “Concepts and Commitments in the Study of
Democracy.” In Problems and Methods in the Study of Politics. Ian
Shapiro, Rogers Smith, and Tarek E. Masoud, eds. Cambridge:
Cambridge University Press, 274–306.
Wedeen, Lisa. 1999. Ambiguities of Domination: Politics, Rhetoric,
and Symbols in Contemporary Syria. Chicago: University of Chi-
cago Press.
Yanow, Dvora. 1997. How Does a Policy Mean? Interpreting Policy
and Organizational Actions. Washington: Georgetown University
Press.
Yanow, Dvora and Peregrine Schwartz-Shea, eds. 2006. Interpreta-
tion and Method: Empirical Research Methods and the Interpre-
tive Turn. Armonk, NY: M.E. Sharpe.
17
Qualitative & Multi-Method Research, Spring 2009
tive from those that can be answered using quantitative meth-
Teaching Interpretive Methods in Political ods, so, too, learning interpretive methods can enlarge stu-
Science: The Challenges of Recognition dents’ capacities for imagining ways to approach research prob-
lems. Methodological pluralism and theoretical pluralism to-
and Legitimacy gether contribute to what Dryzek (1986) has termed “lateral
progress”—the proliferation of distinct research traditions from
Peregrine Schwartz-Shea which societal actors can draw as they encounter historically
University of Utah new problems.
psshea@poli-sci.utah.edu Yet as the opening scenario attests, those teaching inter-
pretive methodologies to graduate students still face two sig-
nificant, intertwined challenges—what, for the purposes of
Recognition – n.
this essay, I call the problem of recognition and the problem of
1. The perception that somebody or something has been
legitimacy. In what follows, I describe how I tackle these twin
seen before or an identification based on such perception;
problems in a required, semester-long research design semi-
2. Appreciation or fame earned by an achievement;
nar. The primary purpose of the course is to introduce stu-
3. Acknowledgement of validity.
dents to elements of research design and to have them prepare
—Encarta World English Dictionary a design on a topic of their choosing. After discussing how I
use the course to deal with these two problems, I present an
Tell a disciplinary colleague that you are teaching quanti-
assignment that concretely illustrates the disciplinary value of
tative or comparative case study methods and you will likely
methodological pluralism. The assignment demonstrates in
get an understanding nod of the head.1 Likewise, graduate
action not only the utility of interpretive methods, but also its
students are reading quantitative research in their seminars
validity as science, the third aspect of recognition noted by
and the term case study will be familiar to all of them no matter
the dictionary definition above.
their subfield. But tell that same colleague or student that you
include interpretive methods in your syllabus and you are likely The Problem of Recognition
to be met with a puzzled look or an outright question: “What’s
All graduate students, and especially those pursuing doc-
that?” Students, too, may be uncertain of the meaning of “in-
torates, likely have encountered research that fits into the cat-
terpretive methods,” asking themselves, “Is that the same as
egory “interpretive,” although such encounters vary greatly
qualitative methods?” and “Have I read anything like that in
by disciplinary subfield and often the research studies do not
my seminars?”
appear under that specific label. International relations and
This lack of both name and content recognition is one
comparative politics students may have read “constructivist”
legacy of the behavioral revolution’s impact on political sci-
research (although not all such research is straightforwardly
ence. With the rise of survey research and the emergence of
interpretive); American politics students may have encoun-
computing capacity, quantitative methods became the meth-
tered Fenno’s (1978) study of Congress or at least his classic
ods widely required of all doctoral students while qualitative
phrase, “soaking and poking”; students of bureaucracy and
methods were offered as optional courses for those specializ-
policy implementation (e.g., in public administration and pub-
ing in comparative politics (Schwartz-Shea 2003, Bennett et al.
lic policy concentrations) may have encountered the classic
2003). There was a concomitant narrowing in the usage of the
studies of Peter Blau (1963), Herbert Kaufman (1960), Michel
term “empirical,” such that many equated it only with “quanti-
Crozier (1964), Graham Allison (1973), or Jeffrey Pressman and
tative” research and, similarly, “methodologist” came to mean
Aaron Wildavsky (1973)—each taking an observational and/
someone specializing in statistical techniques (Schwartz-Shea
or documentary approach characteristic of participant-obser-
and Yanow 2002). In this context, the possibility of interpretive
vation or political ethnography, both of which are common in
methodologies for empirical research was effectively foreclosed.
interpretive research; political theory students are likely to
In the aftermath of the 2000 Perestroika movement (Monroe
have engaged, at a minimum, the idea of a “close reading” of
2005) and with the formation of Consortium on Qualitative
texts, a concept that also falls under this label. And any stu-
Research Methods and the QMMR organized section of APSA,
dent who has taken a broad-ranging philosophy of social sci-
qualitative methods have emerged into the disciplinary sun-
ence course will have heard of the “interpretive turn” (Taylor
light and, now too, as indicated by this symposium, interpre-
1977, Rabinow and Sullivan 1979, 1985, Hiley et al., 1991). (Need-
tive methods and methodologies are beginning to gain a foot-
less to say, these subfield generalities need to be contextualized
hold in graduate curricula.
by departmental variation of the sort documented in Schwartz-
There is now widespread recognition that the lack of plu-
Shea 2005.)
ralism in methods training has been problematic for the disci-
A first task, then, for one teaching a course that incorpo-
pline as a whole. Not only is it clear that no one method is
rates interpretive methodologies and/or methods is explica-
appropriate for all research questions but, also, that a lack of
tion of the label “interpretive”—something that is not as
pluralism in methods training can result in what might be called
straightforward as it might first appear. After all, quantitative
“trained incapacity” (Burke 1965), the inability to ask ques-
scholars interpret their findings, and theorists interpret texts!
tions in different ways. Just as learning qualitative methods
Methods texts do not, on the whole, lend clarity to the situa-
can help students to frame questions in ways that are distinc-
18

tion. As of 2000, the texts marketed to political scientists failed
to include interpretive methods as a possibility for social sci-
entific research (Schwartz-Shea and Yanow 2002). Texts that
do include interpretive methods such as Neuman (2006, 6th
edition) or Creswell (1994) use the term “qualitative.” How-
ever, inspection of their representations and tables (e.g., pages
15 and 5 respectively) show that what these authors mean by
“qualitative” is quite different from the qualitative tradition
that has recently been developed in certain areas of sociology
and political science (the latter being strongly associated with
the comparative politics scholars who founded the QMMR
Organized Section). This means that it is imperative to clarify
for students that they may have encountered interpretive re-
search under the rubric of “qualitative” research in the litera-
ture they are reading outside the discipline, e.g., in interpretive
sociology, social geography, communications studies, cultural
anthropology, etc., or in the European tradition of social theory
(e.g., Foucault 1970, Schütz 1967, etc.), where it means what in
political science, social movement scholarship (e.g., Benford
1997), and other disciplines is being called “interpretive.”
The task of recognition is further complicated by the vast
array of analytic techniques used by interpretive researchers
(from hermeneutics and metaphor analysis to discourse analy-
sis, framing, and narrative and story-telling approaches) and
by the fact that similar terms may be used within interpretive
and qualitative political science—but in very different ways!
For example, interviewing may be conducted for the purposes
of extracting “facts” from interviewees or for understanding
how they make sense of their worlds (although these need not
be contradictory). Similarly, ethnography can be conducted
based on either realist or interpretive ontological and episte-
mological presuppositions. As this last example intimates, ex-
plaining what makes an interpretive approach distinct from
quantitative and qualitative ones requires delving into the realm
of philosophy of social science. Because some students may
not have (yet2) had such a course, I use the grid displayed in
Table 1 as an additional way to explain this distinctiveness
using basic ideas—i.e., theory, data—they are like to have
encountered in their previous seminars.
Column one of Table 1 lists seven components that con-
stitute common ways of talking about research (from purpose
to standards). These provide a set of methods concepts for
comparing two approaches to research, “variables-oriented”
(column two) and “interpretive” (column three). In the second
and third column headings I use the word “gestalt” to empha-
size that a research approach is more than the sum of its parts.
That is, each approach has a holistic, even intuitive “feel”—a
way in which the researcher “approaches” every research
task, from the framing of research questions to assessment of
what constitutes “good” research. In the second column head-
ing, I collapse quantitative and qualitative approaches into a
category I call variables-oriented research, which, through
the strength of the contrast, makes clear that interpretive re-
searchers do not think about their research in terms of vari-
ables—a point that some students find rather startling. (It
also highlights that the distinction has nothing to do with
numbers.) This point comes at the expense of eliding debates
Qualitative & Multi-Method Research, Spring 2009
between qualitative and quantitative researchers over the
value of “large n” and “small n” studies (King, Keohane, and
Verba 1994) and the relative value of variables-oriented and
case-oriented research (Ragin 1997). While the evolution of
these debates is discussed in the seminar, this elision as por-
trayed in the table seems less worrisome in 2009 than it might
have been earlier because considerable consensus seems to
have emerged, at least as judged from the texts by Brady and
Collier (2004) and Gerring (2007), concerning the similarities
between quantitative and qualitative methods (as understood
in contemporary political science). Brady and Collier empha-
size shared standards for assessing quantitative and qualita-
tive studies; Gerring calls quantitative research “cross-case
study” research in order to analyze commonalities as well as
tradeoffs in the choice between cross-case (quantitative) and
case study (qualitative) research designs. Finally, the third
column heading, interpretive research, similarly ignores dis-
agreements among interpretive researchers themselves, such
as (1) whether “causality” should be rejected or re-conceptu-
alized (cf. Hansen 2006, Yanow and Schwartz-Shea 2006); and
(2) epistemological and ontological differences between real-
ist and interpretive IR constructivists (cf. Wendt 1999, Hansen
2006).
Once these caveats have been explained and emphasized,
Table 1 provides a succinct overview that builds on what
students have been learning in their other seminars (or in
quantitative methods courses) in a way that highlights the
interpretive possibility. Although each row could be discussed
in some depth, I will briefly summarize the interpretive col-
umn, as that is less likely to be familiar to readers of the QMMR
Newsletter: (1) Interpretive researchers make human meaning
making central to their analyses and they also conceive of
research purposes as including more than causal reasoning
and generalizable causal laws. (2) Interpretive researchers theo-
rize in a different fashion—most notably, preferring concept
formation that is grounded in the experience and language of
those studied (although what Clifford Geertz, 1973, calls the
“experience-distant” concepts of scholars are also important).
(3) Interpretive research design emphasizes the need for flex-
ibility as researchers investigate and learn about their chosen
topics, and this, in turn, requires a new language of design.
For example, variables-oriented researchers (both quantita-
tive and qualitative) concerned with “sampling” worry about
the extent to which a selected case represents a broader popu-
lation. Interpretive researchers, in contrast, emphasize “expo-
sure” to facets of the evidence and setting as a means of
ferreting out commonalities, differences, and ambiguities in
meaning making. (4) Interpretive researchers respect the
“genre” of the data and do not necessarily prefer to transform
evidence into indicators that can be measured. (5) The spe-
cific analytic techniques chosen by interpretive researchers
depend on the genre of the data—from the vast array of pos-
sibilities for word data to methodological innovations applied
to visual, sound, and spacial data (Bauer and Gaskell 2002,
Harper 2003, Yanow 2006). (6) Interpretive researchers reject
the notion of “causal laws” as inappropriate to a social sci-
ence that emphasizes the historically constitutive nature of
19
Qualitative & Multi-Method Research, Spring 2009

Table 1: Research Components Compared Across Two Research Approaches

Components Gestalt I Gestalt II

Variables-Oriented Research Interpretive Research

Research Purpose Build universal (ahistorical and Focus on meaning making, but also
acultural) theory or mid-level emancipatory/critical and explanatory
theorizing; prediction and control purposes

Theory and Concept A priori, deductive theorizing and Grounded theorizing; concept formation
Formation concept formation in terms of participants’ worldviews

Design Hypothesis and model specification; Flexibility expected as scholar learns; a

selection and measurement of indicators new language of design; e.g., exposure,
not sampling

Data Quantitative preferred as “the best” Respects the genre of the data: numbers,
words, images, space

Analysis Statistical preferred, assuming Depends on the genre of the data

assumptions are met, e.g. unit
homogeneity
Causality Understood as universal laws; Sherlock Holmes, constitutive,
causal mechanisms contingent, configurational

Standards Validity and reliability; internal and Trustworthiness, thick description, tri-
external validity; objectivity angulation, reflexivity, member checks,
audit, negative case analysis; reflexivity

human meaning making—preferring, instead, conceptual-
izations of causality that are grounded in particular cases or
events (as in Sherlock Holmes’ piecing together of concrete
clues); framed as “constitutive” of human understanding and
conduct; or tied closely to context (“configurational”). (7) Be-
cause of all of these differences, interpretive researchers have
developed standards of assessment that are appropriate to
the epistemological and ontological presuppositions and the
purposes of that research gestalt (for an overview of these,
see Schwartz-Shea 2006).
Although Table 1 glosses over many complications (which
I do discuss in class), it provides a “quick and dirty” summary
that can aid students to recognize research studies that are
neither quantitative nor qualitative but are, instead, better un-
derstood under the label of interpretive science.
The Problem of Legitimacy
The possibility of using an interpretive approach to re-
search and research design evokes varied reactions from
graduate students—ranging from excited enthusiasm to skep-
ticism or disinterest (as in, “That’s not relevant in my field or
to my research interests”). Yet even the excited few worry
about the recognition problem, asking: “If I want to use this
approach for my thesis or dissertation research, how will I
convince my committee to approve it?” Of course, all stu-
dents are understandably attuned to the fundamental issues
of career survival. Where could I publish such research?
Would a dissertation using this approach get me a job? Fund-
ing? Tenure? (And, at a deeper level—not always articulated—
is this approach really scientific?)
The substantive value of an interpretive approach is in-
evitably bound up with these sorts of questions. I address
them frankly because students deserve to understand the
contemporary “lay of the land.” There is a reason the recogni-
tion problem exists, and it is bound up with journal publishing
practices, hiring practices, and departmental methods curricula
(Schwartz-Shea 2003). But I also share with them my sense of
how much has changed in the last ten years or so—that an
interpretive community has coalesced within political science;
that I know, personally, young scholars who did interpretive
dissertations and got jobs, in some cases at top research uni-
versities; that some journals are already open to interpretive
approaches and more are opening up, and that good interpre-
tive research is being published; and that interpretive schol-
ars can find support within various sections of APSA (includ-
ing QMMR), WPSA, and, increasingly, MPSA (e.g., its sec-
tions on political sociology and political anthropology), as
well as at ISA and elsewhere.
I emphasize two additional themes during the course of
the semester that help to further assuage these fears. First is
the prominence of the debate around problem-driven research
(Shapiro 2002; Norton 2004) that subordinates method selec-
tion to research question. The second is the widespread dis-
cussion of the value of methodological pluralism.
As an ideal, problem-driven research promotes a vision
of scholars passionately engaged in research about a topic

20

such that any method and any theory will be utilized if it might
shed light on the substantive issues under investigation.
Rather than scholars entrenched in a particular theory or
method (as a mark of their identity), investigators choose
their “tools” flexibly, moving comfortably among statistical,
comparative case study, and interpretive approaches. The
strength of this vision is its emphasis on substance; it is not
that methodological and theoretical debates are unimportant
(for these are the coin of the realm in academic research) but
that such debates should be about something that matters, a
message that graduate students sometimes need to be re-
minded of in the midst of socialization to disciplinary norms.
Although there is much to admire in this ideal of method-
ological pluralism, it is not completely accurate as a descrip-
tion of scientific practices. As the critics of the problem-driven
research contend, “in the world of real research, social scien-
tists do not dream up ‘problems’ to investigate out of thin air,
divorced from concerns of theory and methodology, and only
then search for precisely the right method” (Atkinson et al.
2003: 99). Instead, problem formation is intricately intertwined
with both theoretical and methods training. Here, precisely, is
the value of methodological pluralism enacted in a design
course. Students may be drawn by background or aptitude to
a particular approach (and specialization is likely necessary to
gain depth and expertise) but exposure to more than one ap-
proach helps to develop an appreciation for what I call their
“methodological others”—those persons or projects that use
approaches that they themselves do not use. (And those of
us teaching methods courses need to be careful in handling
multiple methodologies that we don’t in some way, explicit,
subtle or unconscious, christen one particular method as more
“scientific” than others.) This course in research design oc-
curs at a point in students’ careers when they may be able to
develop such an appreciation. The “approaches” assignment,
discussed next, serves as a concrete lesson in pluralism that
students may remember as they go forward in their research
and teaching careers.
The “Approaches” Assignment
Because the research design course is required of all stu-
dents no matter their primary field, it is imperative that I take
into account the methodological traditions that have histori-
cally been associated with the various subfields. Therefore, I
cover three methodological approaches characterized as
“quantitative behavioral,” “comparative case study,” and
“interpretive.”
The first part of the course (The Three Approaches—
Getting Started) introduces these three approaches using pub-
lished exemplars, along with introductory readings from the
assigned texts. This primes students for thinking about the
approaches in relation to their first assignment—providing
an initial two pages on a possible topic for their research
design, which is due at the end of the semester. For the follow-
ing four class sessions (Part II: The Three Approaches—Dig-
ging Deeper), the three approaches are systematically com-
pared in terms of their perspectives on concept formation,
evidence generation, analysis and explanation, and design.
Qualitative & Multi-Method Research, Spring 2009
The approaches assignment which I discuss here is due at the
end of these class sessions, which falls at the midpoint of the
semester. (A second draft of the first research design assign-
ment is also completed during these weeks).
The inspiration for what I call the “approaches” assign-
ment came from an online, three-page, 43-question survey that
faculty were asked by university administrators to complete
on the quality of the campus library. See Table 2 for a partial
representation of the survey. For 27 of the questions, I was
asked to report, using a nine-point Likert scale (“1 is lowest, 9
is highest”), the “minimum level of service that you would find
acceptable,” the “level of service that you personally want,”
and the “the level of service that you believe our library cur-
rently provides” (original emphasis). I found the survey’s de-
sign extremely irritating, even alienating. It was so convoluted,
and its questions were so unrelated to my experiences with
the library that I thought the data produced by it would be
worthless (and I sent an email to that effect to library adminis-
trators). But I also thought that the survey instrument might
be useful as a teaching moment in the research design course
and crafted an assignment around it.
In the assignment, the students’ task is to design research
to assess “library quality.” The advantage of this research
topic is that it is far removed from typical disciplinary ques-
tions, thus promoting the bracketing of theoretical perspec-
tives and disagreements. Additionally, they need not spend
time on a literature review (because that time is better spent
on their individual project designs).
Depending on the number of students in the class, stu-
dents may work as individuals or as teams. Individuals and/or
teams might be assigned to particular approaches or allowed
their own choice—so long as all three approaches are repre-
sented in the oral presentations. Here, I report the result of a
class in which, fortuitously, I had only twelve students. In
that instance, I divided the twelve into two teams of six and
each team further divided itself into three pairs across the
three approaches, thus creating two teams per approach and
also allowing students some leeway to select both partners
and approaches. Students were given the following instruc-
tions:
It is up to each pair to decide what particular methods to
use—being sure to mimic the “spirit” of the approach.
For example, the quantitative behavioral approach could
employ a survey, direct observation, or budget figures
but the emphasis is on precise answers to the question of
quality. A comparative case study approach could em-
ploy observation, interviews, or documentary evidence
but, here, the emphasis is on what can be gained from
comparison among libraries. The interpretive approach
could use participant-observation, interviews, or analy-
sis of documentary evidence but this approach empha-
sizes the meaning making of library users.
It is in the students’ oral presentations that the class as a
whole has the chance to compare how the design of research
varies by approach. On this particular presentation day, ev-
eryone in the class was able to compare within each approach
21
Qualitative & Multi-Method Research, Spring 2009
Table 2: Sample Questions—ARL Survey on Library Quality
When it comes to... My Minimum
Service Level is
(1) Employers who instill Low High
Ο Ο Ο Ο Ο Ο Ο Ο Ο
confidence in users
1 2 3 4 5 6 7 8 9
(2) Making electronic resources Low High
Ο Ο Ο Ο Ο Ο Ο Ο Ο
accessible from my home or office 1 2 3 4 5 6 7 8 9
(3) Library space that inspires Low High
Ο Ο Ο Ο Ο Ο Ο Ο Ο
study and learning
1 2 3 4 5 6 7 8 9
(as we had two teams per approach) as well as across all three
approaches. Additionally, students posted their designs on
WEBCT and I posted comments on each design to further
facilitate cross-approach comparison, assessment, and dis-
cussion.
As might be expected, students studying comparative
politics gravitated to the comparative case study approach
and the teams they formed took advantage of their country
backgrounds—Japan, China, Germany, and the US. They for-
mulated their designs to ask about how libraries are organized
in these different national contexts in order to assess how
cultural, historical, and linguistic differences affect judgments
of quality. Given their previous substantive coursework, as
well as readings from the research design course, teams spent
a lot of time in their proposals on case justification, issues of
unit heterogeneity, and possible causal mechanisms. Given
the class discussion of the problems of the library survey, the
quantitative-behavioral teams spent a lot of time in their pro-
posals and presentations talking about the kinds of surveys
that would generate believable evidence. As befits the spirit of
that approach, they addressed questions of achieving repre-
sentativeness via various forms of sampling. One team de-
cided to use unobtrusive measures to first assess library us-
age. Then, on the basis of this evidence, they targeted their
surveys to those patrons who had recently used those ser-
vices, in this way matching questions of quality to patrons’
specific library experiences.
The two interpretive teams used quite different ap-
proaches. One used a “story telling” methodology to elicit
library users’ concrete experiences of being treated well or
badly by library personnel and to obtain, in the “patron’s own
words,” what mattered most to them about the library. The
other team’s design discussed at length some of the elements
that are distinctive to the interpretive approach—its open-
ended nature, sensitivity to context and political power, and
researcher reflexivity. In doing so, they ended up shifting the
research question in a subtle way or, perhaps more accurately,
they expanded the research question. So instead of asking
22
My Desired Perceived Service
Service Level is Performance is
Low High Low High N/A
Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Low High Low High N/A
Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Low High Low High N/A
Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο Ο
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
the rather abstract question, “What makes a quality library?”,
they asked, “Who gets to define the meaning(s) of a quality
library experience?” When the question is framed in that way,
the political underpinnings of research become much more
apparent. The team included in their design not only users
but also librarians: managers at or near the top of the library
institutional structure but also librarians lower in the hierar-
chy. The potential results from such a research design would
be quite different. Instead of identifying specific factors that
top administrators can use to “improve poor performance” in
the standard bureaucratic way, the study identifies areas of
agreement but also points of conflict and disagreement spe-
cific to the context. Instead of “one best way,” coming down
from those at the top, the research findings could invite more
engagement with the results across institutional hierarchies
and specialties, fostering understanding between librarians
who may serve different sorts of patrons in a variety of ways.
Thus, both the research process and the use of the results
imply a different politics than is typical in other approaches.
At the end of the presentations, there was a lively discus-
sion about the significant differences among the three ap-
proaches. The take-away message was the inevitable, complex
intertwining of our understanding of what research is about
(purpose) with how we frame our question about the topic,
which, in turn, affects the research design and findings. It is
one thing to tell students that these elements interact. It is
much more powerful to have them experience this interaction
by comparing what they have done to what their peers have
done. Each presentation revealed the specific insights afforded
by each approach—helping the students to understand why
someone might use that approach (even if it was not the one
that they preferred).
Toward Methodological Pluralism: Recognition and
Legitimacy of Interpretive Science
By the end of the semester students should be able to
recognize interpretive research when they encounter it and
explain how it differs in general terms from qualitative and
 Qualitative & Multi-Method Research, Spring 2009
quantitative research. Students should also be able to com- 1991. The Interpretive Turn. Ithaca, NY: Cornell University Press.
pare the three approaches in terms of what each approach Kaufman, Herbert. 1960. The Forest Ranger: A Study in Administra-
seeks to accomplish and should appreciate the particular in- tive Behavior. Baltimore, MD: Published for Resources for the
Future by Johns Hopkins Press.
sights each approach can generate. By means of the syllabus
King, Gary, Robert Keohane, and Sidney Verba. 1994. Designing
design and the approaches assignment, an interpretive ap-
Social Inquiry. Princeton: Princeton University Press.
proach becomes legitimized as a possibility within the realm of Monroe, Kristen Renwick, ed. 2005. Perestroika! The Raucous Re-
political science research. And, finally, some students do bellion in Political Science. New Haven: Yale University Press.
choose to frame their research question and formulate its de- Neuman, Lawrence W. 2006. Social Research Methods: Qualitative
sign using an interpretive approach. and Quantitative Approaches. Boston: Allyn & Bacon.
Norton, Anne. 2004. “Political Science as a Vocation.” In Problems
Notes and Methods in the Study of Politics. Ian Shapiro, Rogers M. Smith,
and Tarek E. Masoud, eds. New York: Cambridge University
1
Thanks to Shaun Bowler, David Pion-Berlin, and the faculty at Press, 67–82.
the University of California-Riverside for the opportunity for dia- Pressman, Jeffrey L. and Aaron Wildavsky. 1973. Implementation.
logue about the nature of interpretive research. Thanks, also, to Dvora Berkeley: University of California Press.
Yanow for her references and every-ready editing pen! Rabinow, Paul and William M. Sullivan, eds. 1979. Interpretive So-
2
At the University of Utah, a philosophy of social science course cial Science. Berkeley: University of California Press.
is required of all doctoral students, although they may not have taken Rabinow, Paul and William M. Sullivan, eds. 1985. Interpretive So-
it prior to the design course. Masters students are required to take cial Science, 2nd ed. Berkeley: University of California Press.
design but not philosophy of social science. Ragin, Charles C. 1997. “Turning the Tables: How Case-Oriented
Research Challenges Variable-Oriented Research.” Comparative
References Social Research 16, 27–42.
Schwartz-Shea, Peregrine. 2003. “Is This the Curriculum We Want?
Allison, Graham. 1973. Essence of Decision. Boston: Little, Brown.
Doctoral Requirements and Offerings in Methods and Methodol-
Atkinson, Paul, Amanda Coffey, and Sara Delamont. 2003. Key
ogy.” PS: Political Science and Politics 36, 379–386.
Themes in Qualitative Research: Continuities and Change. Wal-
Schwartz-Shea, Peregrine. 2005. “The Graduate Student Experience:
nut Creek, CA: AltaMira Press.
‘Hegemony’ or Balance in Methodology Training?” In Perestroika!
Bauer, Martin W. and George D. Gaskell, eds. 2002. Qualitative
The Raucous Rebellion in Political Science. Kristen Renwick Mon-
Researching with Text, Image and Sound. London: Sage.
roe, ed. New Haven: Yale University Press, 374–402.
Bennett, Andrew, Aharon Barth, and Ken Rutherford. 2003. “Do We
Schwartz-Shea, Peregrine. 2006. “Judging Quality: Evaluative Crite-
Preach What We Practice? A Survey of Methods in Political Sci-
ria and Epistemic Communities.” In Interpretation and Method:
ence Journals and Curricula.” PS: Political Science and Politics 36,
Empirical Research Methods and the Interpretive Turn. Dvora
373–378.
Yanow and Peregrine Schwartz-Shea, eds. Armonk, NY: M.E.
Benford, Robert D. 1997. “An Insider’s Critique of the Social Move-
Sharpe, 89–113.
ment Framing Perspective.” Sociological Inquiry 67:4, 409–430.
Schwartz-Shea, Peregrine and Dvora Yanow. 2002. “‘Reading’ ‘Meth-
Blau, Peter. 1963 [1953]. The Dynamics of Bureaucracy. Chicago:
ods’ ‘Texts’: How Research Methods Texts Construct Political
University of Chicago Press.
Science.” Political Research Quarterly 55, 457–486.
Brady, Henry E., and David Collier, eds. 2004. Rethinking Social
Schütz, Alfred. 1967. The Phenomenology of the Social World. Chi-
Inquiry: Diverse Tools, Shared Standards. Lanham, MD: Rowman
cago: Northwestern University Press.
and Littlefield.
Shapiro, Ian. 2002. “Problems, Methods, and Theories in the Study
Burke, Kenneth. 1965 [1935]. Permanence and Change. Indianapo-
of Politics, or: What’s Wrong with Political Science and What to
lis, IN: Bobbs-Merrill.
Do About It.” Political Theory 30, 596–619.
Creswell, John W. 1994. Research Design: Qualitative & Quantita-
Taylor, Charles. 1977. “Interpretation and the Sciences of Man.” In
tive Approaches. Thousand Oaks, CA: Sage.
Understanding and Social Inquiry. Fred R. Dallmayr and Thomas
Crozier, Michel. 1964. The Bureaucratic Phenomenon. Chicago: Uni-
A. McCarthy, eds. Notre Dame, IN: University of Notre Dame
versity of Chicago Press.
Press, 101–131.
Dryzek, John. 1986. “The Progress of Political Science.” Journal of
Wendt, Alexander. 1999. Social Theory of International Relations.
Politics 48:2, 301–320.
Cambridge: Cambridge University Press.
Fenno, Richard. 1978. Homestyle: House Members in their Districts.
Yanow, Dvora. 2006. “How Built Spaces Mean: A Semiotics of
Boston: Little, Brown.
Space.” In Interpretation and Method: Empirical Research Meth-
Foucault, Michel. 1970. The Order of Things: An Archaeology of the
ods and the Interpretive Turn. Dvora Yanow and Peregrine Schwartz-
Human Sciences. New York: Random House.
Shea, eds. Armonk, NY: M.E. Sharpe, 349–366.
Geertz, Clifford. 1973. The Interpretation of Cultures. New York:
Yanow, Dvora and Schwartz-Shea, Peregrine. eds. 2006. Interpreta-
Basic Books.
tion and Method: Empirical Research Methods and the Interpretive
Gerring, John. 2007. Case Study Research: Principles and Practices.
Turn. Armonk, NY: M.E. Sharpe.
Cambridge: Cambridge University Press.
Hansen, Lene. 2006. Security as Practice: Discourse Analysis and
the Bosnian War. London: Routledge.
Harper, Douglas. 2003. “Reimagining Visual Methods: Galileo to
Neuromancer.” In Collecting and Interpreting Qualitative Materi-
als. Norman K. Denzin and Yvonna S. Lincoln, eds. Thousand
Oaks, CA: Sage, 176–198.
Hiley, David R., James F. Bohman, and Richard Shusterman, eds.

23
Qualitative & Multi-Method Research, Spring 2009
Teaching Gender and Politics:
Feminist Methods in Political Science
Mona Lena Krook
Washington University in St. Louis
mlkrook@wustl.edu
Feminist research in political science is marked by two
major contributions: (1) introducing the concept of “gender”
and (2) expanding the definition of “politics.” Given its origins
in feminist theory and activism, it is guided by scholarly and
political aims to transform the study and the practice of poli-
tics (cf. Hawkesworth 2006). These commitments enable femi-
nist scholars to identify new research questions, as well as to
approach traditional topics in novel ways, using a variety of
research tools. However, rethinking the content and methods
of political analysis has important implications for how to teach
political science by raising questions about what political sci-
entists study and how and why they study these particular
topics. It also poses certain challenges, or presents new op-
portunities, for political science pedagogy by compelling pro-
fessors to devise innovative techniques for communicating
material and fostering self-reflection among students, who may
resist or embrace central tenets of feminism. To explore how
feminist scholars have met these challenges, this article exam-
ines 45 syllabi for courses on women, gender, and politics
taught at various universities in the United States and West-
ern Europe between 2002 and 2008.1 The analysis begins with
a short introduction to trends in gender and politics research
and interdisciplinary debates on feminist research methods. It
then takes a closer look at the syllabi to illuminate some shared
features of course content, as well as to make note of course
readings, formats, and assignments that reflect feminist com-
mitments to learning and personal transformation. The goal is
to raise awareness of feminist tools and teaching techniques
as a means for assessing their potential contributions for other
areas of political science.
Research on Women, Gender, and Politics
Feminism is often defined as the belief in the social, eco-
nomic, and political equality of women and men. However,
there are in fact a variety of feminist approaches, which differ
in terms of how they conceptualize and seek to alter the status
quo.2 Liberal feminists focus mainly on equality, seeking to
gain rights for women that are already guaranteed to men.
They argue that achieving concrete gains requires engaging
with formal politics. They contend that although this sphere
has traditionally been dominated by men, there is nothing in-
herent about this domination. For this reason, they anticipate
that as more women enter the public realm, the gendered na-
ture of politics and public policy can be overcome to create
equality for all. Radical feminists, in contrast, emphasize differ-
ence, aiming to focus on and value women as women, rather
than as individuals who aspire to a male standard. As such,
they are more skeptical about the value of participating in “poli-
24
tics as usual,” which they argue is inherently patriarchal and
thus could never be employed to pursue feminist ends. They
prefer political strategies that revalue the feminine, foster soli-
darity among women, and raise awareness of women’s experi-
ences through collective consciousness-raising. Postmodern
feminists are also interested in difference, but focus more on
how categories like “women” and “men” are represented
through discourse. Theorizing the fluid and relational aspects
of identity and experience, they stress contradictions and mul-
tiplicities in definitions of “women” and “women’s issues.”
While this approach avoids the charges of essentialism that
have been directed towards liberal and radical feminism, it also
has the effect of undermining the prospects for mobilizing by
women as women for social, economic, and political change.
Feminism thus poses varied challenges to existing modes
of political analysis. All the same, the research that falls under
this rubric shares roughly similar goals to incorporate gender,
expand politics, and promote change. The concept of “gen-
der” is often considered the key contribution of feminism as
an intellectual and political project. Although often elided with
“women” in popular and scholarly discourse, feminists are
careful to distinguish between “sex,” biological differences
between women and men, and “gender,” the social meanings
given to these distinctions. A shift to gender has two broad
implications for political research: (1) it moves the analytical
focus away from biological sex, which treats men and women
as binary opposites, to constructed gender identities, which
view masculinity and femininity as features that exist along a
continuum, often in combination with other identities, and (2)
it replaces exclusive concern with women in politics and pub-
lic policy with attention to the impact of masculinities and
femininities, as well as relations between men and women, on
political inputs and outcomes (Krook and Childs 2009). Given
women’s ongoing exclusion, focusing on “women” remains
crucial for mapping patterns of political access, behaviors,
and effects. However, theories of gender offer a chance to
delve more deeply into these dynamics by bringing men into
the analysis as well, thereby making the subject of investiga-
tion the role of masculinities and femininities, and the relative
status of men and women, in the conduct of political life.
A second concern of feminists is to broaden existing defi-
nitions of what is meant by “politics.” Political scientists tend
to use this term to refer to formal processes and institutions of
government and elections. Women’s movement activism in
recent decades, however, has inspired feminists to theorize at
least two additional meanings. One group expands its range
to encompass informal politics and the dynamics of everyday
life. Some scholars insist, for example, that social movements
are a form of political participation on par with engagement
inside the state (Beckwith 2005). At the same time, others
draw attention to the power dynamics that permeate all levels
of social life, including relations within the private sphere of
home and family. Echoing the slogan of second wave femi-
nism, they argue that “the personal is political” (Okin 1979). A
second group, together with postmodern theorists, has
adopted a notion of “politics” as any instance or manifesta-
tion of power relations (Butler 1990). As a result, they are

interested not only in the politics of the state and the politics
of social movements, but also the politics of language, the
politics of exchange, and the politics of representation, which
they have analyzed using a wide variety of research tools.
Both of these feminist innovations have come under chal-
lenge in recent years. On the one hand, there has been in-
creased recognition of the ways in which multiple facets of
identity may interact to shape not only personal interactions
but also large-scale political outcomes. In these debates, schol-
ars have offered various schemes for analyzing how the dy-
namics of gender shape and are shaped by other patterns of
inequality based on race, class, sexuality, ability, and other
features (Hancock 2007; Weldon 2006). On the other hand,
increased globalization, combined with decentralization, has
posed major challenges to traditional configurations of politi-
cal organization. As a consequence, “politics” is now an even
more diffuse entity, with new and developing arrangements
that are not yet well understood (Krook and Childs 2009).
Understanding both sets of trends is crucially important for
the third main element of feminist research: a commitment to
political change. Although feminists of various types espouse
diverse goals, they converge on the opinion that research
should contribute to some type of positive transformation,
whether this entails the broad empowerment of women as a
group or the deconstruction of gendered categories in poli-
tics and public policy.
Perspectives on Feminist Research Methods
The political goals of feminism, while central to the ethos
of this line of research, have been used on occasion as an
argument against feminist scholarship on the grounds that it
fails to be “objective,” as such motives interfere with the dis-
covery of “truth” (Hammersley and Gomm 1997). In response
to such critiques, feminist epistemologists argue for recogniz-
ing the situated and partial nature of all knowledge claims.
Yet, there are ongoing debates as to whether there are specifi-
cally feminist research methods, ranging from broad accep-
tance of existing tools, a position known as “feminist empiri-
cism” (Harding 1986), to various attempts to explore and de-
vise new methods of analysis (Hesse-Biber et al. 2007). Al-
though qualitative methods are often associated more closely
with feminist research, some insist that feminist work can and
should utilize both quantitative and qualitative techniques
(Jayaratne and Stewart 1991; Oakley 1998). This has led schol-
ars to suggest that there are no feminist methods, but that
there is one feminist methodology (Reinharz 1992). Said in
another way, feminists may employ many different research
techniques in their quest for evidence, but share an approach
to collecting and evaluating this evidence. This methodology
is said to be “distinctive to the extent that it is shaped by
feminist theory, politics, and ethics and is grounded in women’s
experience” (Ramazanoglu with Holland 2002: 16). A survey of
feminist methods texts indicates that it comprises four main
elements: paying attention to “gender,” challenging norms of
objectivity and incorporating subjectivity into research, try-
ing to avoid exploiting women as subjects and objects of
knowledge, and empowering women in various ways through
Qualitative & Multi-Method Research, Spring 2009
research (Acker, Barry, and Esseveld 1996; Fonow and Cook
2005; Harding 1987).
As these goals are sufficiently broad, many different re-
search techniques may be employed in a manner consistent
with feminist values. At the same time, some of these con-
cerns overlap with other approaches to social investigation,
especially those that seek to question existing “truths.” What
make feminist studies distinctive are their efforts to adapt
many of the same methods as other researchers in ways that
make them more consistent with feminist concerns. In the use
of interview techniques, for example, feminists have often been
careful to involve research subjects in the construction of
data about their own lives. In the process, they have become
conscious of particular challenges inherent in generating femi-
nist insights—or simply remaining consistent with feminist
goals—when interviewing across age, race, class, gender, and
political differences. Feminists have also discussed ways in
which other techniques may be employed to feminist ends:
archival research—or even starting a group’s own archive—
can help promote knowledge of many different kinds of
women; internet research can reach women who are geographi-
cally dispersed but “virtually” connected in order to study
how they share knowledge or mobilize collectively; content
analysis can provide insights into discursive and visual rep-
resentations of gender through non-traditional research ma-
terials like artwork and other cultural artifacts; and surveys
and statistical analyses can reveal that gender inequalities do
in fact exist, affording crucial leverage for feminist activists in
their efforts to influence public policy (cf. Fonow and Cook
2005; Gottfried 1996).
In other cases, feminist researchers create new methods
in the pursuit of better knowledge of gender relations. The
quintessential method of this type is consciousness-raising,
a crucial tool in second-wave feminism, which typically in-
volves small groups of individuals who meet to discuss their
personal experiences. These gatherings, which may also take
the form of “speak-outs” and “write-ins,” help participants
recognize the hidden and taken-for-granted aspects of their
lives that enable personal transformation and provide insights
for devising strategies for change. Other techniques have
been invented in the course of asking questions whose an-
swers are difficult to access through traditional methods.
These include dramatization through role play, which allows
research subjects to collaborate in research and to find their
own voice; conversation, which presents multiple voices as a
way of gathering and displaying data; identification, which
“breathes life” into the person being studied through the per-
sonal reflections of the scholar doing the study; structured
conceptualization, which entails synthesizing information in
the form of a map in order to display how ideas are related to
one another; photography, which compiles images of the re-
search subjects to tell a visual story of their lives and experi-
ences, sometimes involving their participation in the presen-
tation of findings; and taped self-interviews, which enable
respondents to answer questions at their convenience in the
privacy of their own homes (for a list of details and examples,
see Reinharz 1992). These solutions, combined with extensive
25
Qualitative & Multi-Method Research, Spring 2009
feminist adaptations of existing techniques, suggest—con-
trary to the conventional wisdom—that there may in fact be a
number of feminist research methods, consistent with differ-
ent definitions of “feminism” and various feminist goals in the
research process.
Feminist Methods in Political Science
Feminist discussions of methods and methodologies have
implications for the substance and goals of political research
done in a feminist vein. These, in turn, have ramifications for
how scholars teach political science, in relation to the content
and the pedagogical techniques used in courses on women,
gender, and politics. The 45 syllabi examined for this article
cover a range of topics in American politics, comparative poli-
tics, and international relations. They were collected from pub-
lic and private institutions, undergraduate and graduate
courses, and male and female professors. Attention was paid
to course descriptions, readings, formats, and assignments to
discern whether, how, and to what extent feminist ideas and
concerns were integrated into course material. This review
reveals that instructors, as reflected in the syllabi, did seek to
introduce students to new ways of understanding the politi-
cal world through gendered lenses, often drawing on work
using a variety of different research methods. More interest-
ingly, however, many courses also utilized one or more inno-
vative teaching techniques to (1) engage students with de-
velopments in the “real world” and (2) equip them with new
skills and insights to facilitate personal transformation and
empowerment.
Course Content
An analysis of the syllabi indicates that professors in all
courses endeavor to introduce students to new ways of “do-
ing” political science. Consistent with feminist literature in
the field more generally, they familiarize students with the
concept of “gender,” enlarge the scope of what is considered
to be “political,” and offer insights into possible strategies for
political “change.” Many of these courses begin with a sec-
tion on feminism and gender theory. Providing an overview of
diverse feminist approaches, they emphasize “gender” as a
social construct that has a range of important implications for
political life, starting with the tendency to associate men with
the public sphere and women with the private. Many U.S.-
based courses, in particular, also make conscious efforts to
recognize diversity among women by incorporating attention
to race, class, and sexuality. A key aim is to bring a gender
perspective into an array of topics in political science, explor-
ing the limits of existing paradigms and literatures. In many
cases, this entails calling attention to women as subjects and
objects of public policy, noting the various roles that women
have played as political actors and the often distinct impact
that otherwise neutral-sounding policies may have on women
versus men. A large number of courses address these dynam-
ics by highlighting gaps between women and men in terms of
their voting and legislative behavior. Some also focus on
women’s experiences as a means to explore how women may
draw on norms of femininity in unanticipated ways, for ex-
26
ample by pointing to their skills as mothers to assert their
suitability for leadership roles. The intent in all instances is to
analyze how gender operates in the political realm and, as a
consequence, rethink and reconceptualize political concepts
through a feminist lens.
A second feature of these courses is a shared effort to
address, but also go beyond, the confines of formal politics.
Seeking to break down the public/private divide, many in-
structors draw attention to the partial nature of how “politics”
has been studied, in some cases by quoting the feminist slo-
gan: “the personal is political.” To this end, they point to the
neglected experiences and arguments of women to expand
the range of actors and issues understood to be relevant to
political debate. Nearly all courses include a section on
women’s social movement organizing and participation in other
voluntary sector activities. A key reason for this is historical:
women have largely been excluded from other arenas of politi-
cal participation, like elections, political office, and interna-
tional politics. In addition, while formal barriers, like the lack
of suffrage, have been overcome in most countries, women
still occupy relatively few top-level political positions. As a
consequence, a great deal of research on women and politics
has focused on women’s activities in civil society. Most
courses, however, also address a range of different types of
formal political participation, examining trends in women as
voters, party activists, candidates, and elected officials. Fur-
ther, a sizeable number of courses extend the realm of “poli-
tics” to the media, the judiciary, bureaucratic agencies of the
state, and various types of international organizations. In a
similar set of moves, many courses cover a range of issues
often associated with women, like equality law, educational
equity, workplace and family issues, health, reproductive rights,
and violence against women. However, at the same time, there
is also an increasing tendency to include issues not tradition-
ally viewed through a gender lens, like human rights, devel-
opment, trade, migration, nationalism, national security, war,
and science and technology.
A third notable element of course content is discussion
of efforts to promote political change. One syllabus, indeed,
describes the course in question as “subversive and recon-
structive” and frames it explicitly as an “intellectual and politi-
cal journey.” The aim, according to the instructor, is not only
to expose the limits of conventional modes of political analy-
sis, but also to move from describing the world to thinking
about how to reconstitute these realities. This goal is at-
tempted via several distinct course designs. The majority of
courses focus on politics in a single country, like the U.S., but
many also include some degree of cross-national comparison
as a means to (1) raise students’ awareness of distinct trends
elsewhere and (2) explore why change has occurred in some
countries but not in others. While crucial for improved knowl-
edge of political processes, exposure to such information is
also intended to foster the ability to imagine an alternative to
the status quo. Another technique involves introducing stu-
dents to particular strategies that have been developed around
the globe to empower women and create more gender-sensi-
tive public policy. Instructors in these courses include units

on gender quotas, policies to increase the selection of female
candidates to political office, and gender mainstreaming, an
approach to policy-making that requires considering the
gendered implications of all public policies. They also draw
attention to actors who have played important roles in ad-
vancing gender equality, both expected, like movement activ-
ists, and less expected, like state bureaucrats. A final approach
is to brainstorm and introduce scholarly evidence as to how
political life might change, or not, with the greater inclusion of
women. Discussions highlight the stakes to maintaining the
status quo, but also encourage students to consider possible
limits to change, for example by noting that the increased
presence of women is not always associated with dramatic
shifts in policy outcomes.
Course Methodology
Paralleling debates on feminist research methods, the read-
ings used in gender and politics courses reflect a range of
different methods and methodologies. While none of the syl-
labi surveyed assign books or articles on research methods,
or even address questions related to the philosophy of sci-
ence in their reading lists, almost all include a session or more
on the concepts of “sex” and “gender” and an overview of
different types of “feminism.” Many instructors take care to
emphasize diversity across feminist approaches, noting that
these present distinct frameworks for understanding and ana-
lyzing dynamics of gender, politics, and change. As such,
many syllabi trace the development of feminist thinking, as
well as outline ongoing feminist debates, on a particular topic.
The result is that courses tend to offer insights into the dis-
tinctiveness of feminist analysis, at the same time that they
recognize the multiplicity of feminist contributions to political
science.
In terms of the more specific methods employed in the
readings, what is striking about these courses is their open-
ness to studies using a variety of different research tools.
Although some textbooks are assigned, including Conway,
Ahern, and Steuernagel (2005), Harrison (2003), and Paxton
and Hughes (2007), the vast majority of readings are taken
from articles and book chapters (but see Krook and Childs
2009). In general, the choice of methods in each text is related
to the topic under investigation: archival analyses and eth-
nographies prevail in studies of women’s movements, elite
interviews and statistical analyses of cross-national data when
the subject is women in parliaments, surveys when the ques-
tion relates to elite and public opinion, and textual analysis in
work on law and public policy. In many cases, several meth-
ods are used in conjunction with one another. These patterns
may appear surprising to some, given that feminist research is
often seen as having a preference for qualitative methods.
This is due to perceived difficulties in operationalizing “gen-
der” as a variable for quantitative analysis. While “sex” can
be recorded as a simple dichotomous measure, such an asser-
tion is controversial among some feminists, who argue that
“sex” too is socially constructed, and as such, should also be
understood as existing along a continuum of identity (cf. But-
ler 1990).
Qualitative & Multi-Method Research, Spring 2009
Yet, perhaps one of the most notable features of feminist
work is its “problem-driven” nature: a recent review discovers
a distinctive willingness on the part of feminists to employ
various theoretical frames and to explore possibilities for syn-
thesizing or juxtaposing methods in innovative ways (Krook
and Squires 2006). As such, the diversity of methods employed
in the readings assigned in gender and politics courses simply
provides a reflection of the eclectic tools that feminist research-
ers have used in their pursuit of better knowledge of the politi-
cal world. All the same, there are several methods which ap-
pear to be less prevalent across these courses as a whole,
including game theory and rational choice, qualitative com-
parative analysis (QCA), and interpretive methods. These tools
have been used in gender and politics research, but tend to be
less common. Thus, it is not surprising that such readings are
rarely assigned in these courses. However, it is crucial to em-
phasize that these patterns do not necessarily stem from their
ontological and epistemological incompatibilities with femi-
nist research. Rather, the tendency to use some methods more
than others is more likely connected to the fact that there is
still much to be explored with regard to the gendered nature of
political life.
Finally, it is worth noting that some of the materials as-
signed in the courses reviewed do not simply rely on tradi-
tional “readings.” Some instructors ask students to read nov-
els and watch films for later discussion in class, primarily—it is
assumed—as a means to capture the complexities of women’s
(and sometimes men’s) lived experiences. A large number of
courses also require students to consult and reflect upon a
range of different primary sources. These include, most com-
monly, internet links to online materials like reports from inter-
national organizations; websites of feminist and human rights
NGOs; and databases on policies and statistics related to
women, gender, and politics. Depending on the focus of the
course in question, they may also entail newspaper articles
and opinion pieces on current events; interviews transcribed
in a book or available recorded online; court cases and deci-
sions; texts in political theory, especially in relation to ques-
tions of political representation; and political biographies and
autobiographies. Assigning materials such as these requires
students to offer their own analyses and interpretations of
gendered political events.
Course Pedagogy
The intellectual and political commitments of feminism, in
turn, have an impact on course instruction. In some instances,
a feminist philosophy is explicitly spelled out, as with one
professor who writes: “As a feminist teacher, I am committed
to a mode of learning that is shared and collaborative.” It is the
case that most courses do make use of conventional teaching
formats, like giving lectures, showing films, inviting guest
speakers, holding seminars, and conducting email discussions.
They also have assignments similar to those in other political
science courses, like take-home and in-class exams, research
papers, book reviews, and short reflection papers on course
readings. Nonetheless, the syllabi analyzed here reveal that
many instructors also incorporate more unusual teaching for-
27
Qualitative & Multi-Method Research, Spring 2009
mats and course assignments. Reflecting feminist interest in
political change and empowerment, several take specific steps
to connect the materials introduced inside the classroom to
the “real world.” This involves providing students with op-
portunities to literally go outside the classroom by offering
information about internships, arranging meetings with con-
gressional staffers, watching politics “in action” by attending
a local party women’s conference, and engaging in “service
learning” by tutoring local refugee women. It also entails trans-
forming the outside world into the classroom by inviting vari-
ous political women in to discuss their experiences and paths
to political office. A related strategy is to require students to
bring in news articles for class discussion, using these new
items to link the theories and concepts introduced in class to
help students better understand and analyze recent political
developments.
A second major trend, as indicated in the syllabi, is to use
course assignments as a means for cultivating new skills and
encouraging personal transformation. This is accomplished
through a series of diverse and original course activities and
requirements. Several courses include training sessions on
how to run for elected office. In some cases, this entails invit-
ing guest speakers who offer advice on how to get involved
in politics as activists, campaign workers, and candidates. In
others, students learn specific political skills, like doing back-
ground research, preparing speeches, making presentations,
training in giving television and radio interviews, writing op-
ed pieces and blog entries, and engaging in effective net-
working. Other courses focus on improving students’ com-
munication skills by scheduling tasks that require them to
synthesize and articulate arguments related to gender and
politics. In some instances, the assignment is oriented toward
the collective: several instructors plan in-class debates, either
formal tête-a-têtes or more informal group discussions, for
which students receive some background readings and ori-
enting questions to prepare. In some courses, the assignment
is more individually based: students are asked to make oral
presentations that summarize the readings as a means for ini-
tiating and leading a class discussion, or—perhaps more
dauntingly—to offer a five-minute speech to the class about
the importance of women’s political participation, representa-
tion, and leadership.
Other assignments are more writing-based. One instruc-
tor requires students to write a short advocacy paper making
a case for a specific government policy on an issue affecting
women. A variation used by other professors is to ask stu-
dents to write an op-ed or letter to the editor on an issue of the
students’ choice, with the intent to influence policy-making.
Another project assigned in several courses is elite interview-
ing. One course calls for students to interview a woman active
in influencing policy on women, presumably with the goal of
helping students understand how policy-making “works” and
what the constraints and opportunities are for women to act
on behalf of women as a group. Another instructor requires
students to conduct separate interviews with one man and
one woman in the same leadership position. Students are then
to write a paper reflecting on if and how men and women lead
28
differently in these positions. By making the question of “dif-
ference” an empirical question, rather than a theoretical given,
this assignment encourages students to grapple with the con-
cept of “gender” in a real world laboratory, pushing them to
consider when and where it may be relevant—or not.
A final example is one of the most common assignments
across the syllabi surveyed. At its most basic, this task in-
volves doing research and analyzing the profiles and experi-
ences of individual female politicians. It appears in a number
of different versions. In some courses, students are asked to
write a review of a book, selected from a list of biographies
and autobiographies of political women. Another professor
requires students to keep a weekly journal of a female leader,
living or dead. If the woman is still alive, students are to follow
current events, keep clippings, and discuss successes and
failures. If the figure is historical, students need to find and
analyze as many original sources as possible. A second varia-
tion of this assignment is to write a background paper on a
female member of Congress, focusing on her professional
background, personal history, issue priorities, and committee
allocations. A closely related alternative is to construct a bi-
ography of a female leader, anywhere around the world, which
discusses her upbringing, her rise to power and the context in
which it occurred, and some of the important events and deci-
sions she made during her tenure in office. In some instances,
students are obliged to draw explicit links between these find-
ings and various themes introduced in the course. In others,
they are required to follow and analyze ongoing election cam-
paigns. One instructor asks students to focus on a competi-
tive House or Senate campaign involving a female candidate,
comparing the strategies and behavior of the woman and her
male opponent. Another directs students to use a combina-
tion of academic readings and newspaper reports to develop
a profile of the candidate and evaluate her race in relation to
the literature on female candidates. Taken together, these as-
signments seek to give students a greater understanding of
how women might attain political office, as well as what types
of barriers to political parity still remain.
Conclusions
Feminist research and teaching in political science is thus
marked by efforts to produce better knowledge of the political
world, as well as to engage in a broader project of political
transformation at both individual and collective levels. Some
instructors explicitly recognize that such goals may not be
easily achieved. Indeed, one even acknowledges on the first
page of the syllabus that this may be a “controversial and
even painful course” for some students, in that it is likely to
challenge their prevailing views of the world around them.
Although feminist aims are themselves diverse, the review of
syllabi undertaken here indicates that there are perceptible
trends in how courses on women, gender, and politics are
designed to reflect the intellectual and political goals of aca-
demic feminism. This includes a distinctive willingness to uti-
lize a range of different research methods, in pursuit of an-
swers to different types of questions in feminist political sci-
ence. Yet, the attention given in these courses to personal

transformation and empowerment, in particular, is notable for
its attempts to bridge scholarly writings and political devel-
opments outside the classroom. Together with the feminist
stance on the need to engage in “problem-driven” research,
this approach to teaching and learning may offer novel les-
sons for other scholars, who struggle with how to make poli-
tics “real” and “relevant” for their students. Features of femi-
nist research and teaching might thus be understood as a
model of “good practice,” instructive for many other courses
offered in political science.
Notes
1
Many thanks to Amy Mazur and Gary Goertz for compiling
this collection of syllabi.
2
There are many ways to categorize various feminist approaches.
This article discusses only three, but other variants include socialist
feminism, maternal feminism, and Black feminism, to name but a few.
References
Acker, Joan, Kate Barry, and Joke Esseveld. 1996. “Objectivity and
Truth: Problems in Doing Feminist Research.” In Feminism and
Social Change: Bridging Theory and Practice. Heidi Gottfried, ed.
Urbana: University of Illinois Press, 60–87.
Beckwith, Karen. 2005. “The Comparative Politics of Women’s
Movements.” Perspectives on Politics 3:3, 583–596.
Butler, Judith. 1990. Gender Trouble. New York: Routledge.
Conway, Margaret, David Ahern, and Gertrude Steuernagel. 2005.
Women and Public Policy: A Revolution in Progress, 3rd. ed. Wash-
ington: Congressional Quarterly Press.
Fonow, Mary Margaret and Judith A. Cook. 2005. “Feminist Meth-
odology: New Applications in the Academy and Public Policy.”
Signs 30:4, 2211–2236.
Gottfried, Heidi, ed. 1996. Feminism and Social Change: Bridging
Theory and Practice. Urbana: University of Illinois Press.
Hammersley, Martyn and Roger Gomm. 1997. “Bias in Social Re-
search.” Sociological Research Online 2. http://ideas.repec.org/a/
sro/srosro/1996-53-1.html (accessed February 21, 2009).
Symposium: Conceptualizing and Measuring Ethnic Identity
Measuring the State’s Institutionalization
of Ethnic Categories
across Time and Space
Evan S. Lieberman
Princeton University
esl@princeton.edu
Prerna Singh
Princeton University
prernas@princeton.edu
In this article, we describe a research strategy for measur-
ing the degree of state institutionalization of ethnic categories
across time and space, and we present some preliminary data
associated with this work. Our approach is both qualitative
Qualitative & Multi-Method Research, Spring 2009
Hancock, Ange-Marie. 2007. “When Multiplication Doesn’t Equal
Quick Addition: Examining Intersectionality as a Research Para-
digm.” Perspectives on Politics 5:1, 63–79.
Harding, Sandra. 1986. The Science Question in Feminism. Ithaca:
Cornell University Press.
Harding, Sandra, ed. 1987. Feminism and Methodology: Social Sci-
ence Issues. Bloomington: Indiana University Press.
Harrison, Brigid C. 2003. Women in American Politics: An Introduc-
tion. Belmont, CA: Wadsworth.
Hawkesworth, Mary. 2006. Feminist Inquiry: From Political Con-
viction to Methodological Innovation. New Brunswick: Rutgers
University Press.
Hesse-Biber, Sharlene Nagy, and Patricia Lina Leavy, eds. 2007. Femi-
nist Research Practice: A Primer. Thousand Oaks, CA: Sage.
Jayaratne, Toby Epstein and Abigail J. Stewart. 1991. “Quantitative
and Qualitative Methods in the Social Sciences: Current Feminist
Issues and Practical Strategies.” In Beyond Methodology: Feminist
Scholarship as Lived Research. Mary Margaret Fonow and Judith
A. Cook, eds. Bloomington: Indiana University Press, 85–106.
Krook, Mona Lena and Judith Squires. 2006. “Gender Quotas in
British Politics: Multiple Approaches and Methods in Feminist
Research.” British Politics 1:1, 44–66.
Krook, Mona Lena and Sarah Childs, eds. 2009. Women, Gender, and
Politics: A Reader. New York: Oxford University Press.
Oakley, Ann. 1998. “Gender, Methodology and People’s Ways of
Knowing: Some Problems with Feminism and the Paradigm De-
bate in Social Science.” Sociology 32:4, 707–731.
Okin, Susan Moller. 1979. Women in Western Political Thought.
Princeton: Princeton University Press.
Paxton, Pamela and Melanie M. Hughes. 2007. Women, Politics, and
Power: A Global Perspective. Thousand Oaks, CA: Pine Forge
Press.
Ramazanoglu, Caroline with Janet Holland. 2002. Feminist Method-
ology: Challenges and Choices. Thousand Oaks, CA: Sage.
Reinharz, Shulamit. 1992. Feminist Methods in Social Research. New
York: Oxford University Press.
Weldon, S. Laurel. 2006. “The Structure of Intersectionality: A Com-
parative Politics of Gender.” Politics & Gender 2:2, 235–248.
and quantitative: we attempt to unearth and to carefully clas-
sify key historical facts and texts, and also to develop cali-
brated numerical indices. Because our work is ongoing, we
hope that this article may stimulate suggestions for revisions
to our approach to conceptualization and measurement.
We take seriously the currently dominant constructivist
theoretical understanding of ethnic identification, recognizing
the political activation of such identities as a phenomenon to
be explained with reference to discrete historical events and
processes, not merely as “givens.” This theoretical orienta-
tion has prevailed in large part because of the outstanding
detailed studies by political scientists, anthropologists, soci-
ologists, and historians who offered theoretically compelling
accounts of the ways in which state institutions have been
central to the development of ethnic identities (Anderson 1996,
Gellner 1983, Barth 1969, Laitin 1986, Marx 1998). Meanwhile,
on a quite distinctive track, a large body of research pioneered
by economists, but also vigorously pursued by political scien-
29
Qualitative & Multi-Method Research, Spring 2009
tists, has been concerned with estimating the consequences
of ethnic diversity for a range of substantively important out-
comes, including rates of economic growth, government policy,
and patterns of violence (e.g., Easterly and Levine 1997, Fearon
and Laitin 2003). However, as has been widely noted (see, for
example, Chandra 2001)—and we will not repeat such critiques
here—the quantitative analyses associated with this latter
body of research have rested almost entirely on data that re-
flect little incorporation of constructivist insights. And yet,
the compelling nature of that research program has led many
scholars, including several of the contributors to this sympo-
sium, to attempt to develop new approaches to concept-
ualization and measurement, with the hopes that we may yield
more valid assessments of the causes and consequences of
this ethnic identification and mobilization.
We believe that the unique contribution of the project
described will be to systematically incorporate the insights
from earlier scholarship on state institutions in a manner that
allows us to carry out broader comparative analyses across
time and space. Although we certainly recognize that state
institutions are not the only relevant factors in the construc-
tion of ethnic identities, we believe that a number of studies
have compellingly demonstrated the key role played by these
institutions, warranting this type of expanded investigation.
Our goal is to investigate the history of state institutions on a
country-by-country basis, generating a rich database of both
specific historical facts and the development of an Institution-
alized Ethnicity Index (IEI), which summarizes the degree to
which ethnic distinctions figure within the state-institutional
environment. This is likely to facilitate cross-national research
on the causes and consequences of particular ethnic institu-
tions,1 their aggregates, and their legacies, and should serve
as a useful complement and corrective to extant analyses.
State Institutions that Create or Reify Ethnic Categories
In line with constructivist theories of ethnic group forma-
tion, institutions may create incentives/facilities for, or con-
straints upon, ethnic identification and mobilization. In a
Gramscian manner, institutions may make particular sets of
identities hegemonic (Laitin 1986). They are also likely to re-
flect existing ethnic schemas. We proceed with the core as-
sumption that the higher the frequency with which state insti-
tutions make ethnic distinctions, the greater the likelihood that
both ordinary citizens as well as political elites will perceive
ethnic identity to be a salient category for politics, and press
claims with respect to such identities.
As a first cut at systematizing an approach, we focus ex-
clusively on formal state institutions at the national level. There
exists a high degree of institutional isomorphism in the inter-
national state system, and a well-established battery of instru-
ments available to constitutional planners and to national and
ethnic political entrepreneurs for the identification and cat-
egorization of their populations. We consider a range of key
institutions that are used by most states as organizational and
administrative tools, and we explore the extent to which ethnic
group distinctions are incorporated into such rules. Future
research will need to consider informal and sub-national insti-
30
tutions.
In our conceptualization of institutions, we have at-
tempted to strip away normative concerns of “fairness,” and
the potential effects of institutions (such as violence, socio-
economic disparities, or identification), because these are the
very factors that we hypothesize may cause or be caused by
the state’s institutionalization of ethnic categories. As com-
pared with other approaches, issues of subjective coder in-
terpretation and reactivity are significantly reduced because
we can rely upon written historical documents and we are
able to develop fairly straightforward rules for evaluating the
presence or absence of particular institutions.
We consider only state institutions that have the poten-
tial to directly address questions of identity, ethnicity, and
membership in the national community. Because states seek to
monitor their populations and regulate questions of citizen-
ship, justice, and access to power, they must choose to either
recognize ethnic groups or to ignore them. We do not consider
institutions that indirectly influence ethnic politics in the form
of electoral design, which might have intended and unintended
consequences with respect to the mobilization and organiza-
tion of ethnic groups. We also do not consider in this frame-
work civil society organizations or political parties that en-
gage the state, but which we consider as residing within the
societal domain. As shown in Table 1, we identify nine differ-
ent categories of institutions that have been used by states to
regulate human behavior, often in terms of ethnic categories.
In the sections that follow, we delineate the particular ways in
which state institutions make ethnic distinctions, which serve
as the basis for our coding of the historical record. We also
note our theoretical expectations for how the institution might
provide (dis)incentives for ethnic differentiation, mobilization,
and conflict.
Counting and Identifying:
Censuses and Passports/Identity Cards
Virtually all modern states use a variety of techniques to
count and to identify their citizens. For our purposes, a central
question is the extent to which, in addition to identifying na-
tional citizenship or membership, these institutions also dis-
tinguish among various sub-national ethnic categories. Sev-
eral scholars have already convincingly described the ways in
which such categorizations both reflect and reproduce those
categories as the basis for political competition (Nobles 2000;
Kertzer and Arel 2002). Our fundamental premise here is that in
order for citizens to believe that ethnic-based injustices exist
or to consider mobilization of grievances along these lines,
there must be some information about those groups, and group
members must be identifiable. While other instruments are ob-
viously available for such tasks, the census—both the act of
carrying it out and the data it generates—and identity docu-
ments that label citizens are particularly important.
Recent scholarship emphasizes that the counting of eth-
nic groups is by no means a neutral exercise. The act of enu-
merating ethnicity can itself shape the categories used in eth-
nic classification within society. Scholars claim that official
census categories not only reflect national understandings of
 Qualitative & Multi-Method Research, Spring 2009
Table 1: Ethnic Institutions
Institution Institution Evidence of Institutionalization Theoretical Prediction
Type of Ethnic Categories

Counting and 1. Census Any mention of ethnic categories or labels on Forces individuals to “choose”
Identifying questionnaire or enumeration form. an official identity, or to accept
one that has been assigned;
2. ID Cards/Passports Any mention of ethnic categories or labels on
generates official data in ethnic
documents.
terms, which in turn are used
for ethnic political competition.
Politics and 3. Delegation of There is any legal provision for separate laws Any distinctions in the political
Authority Autonomy or authority for ethnic groups. arena made in terms of access
to the vote or to leadership
4. Voting and Civic The government uses ethnic identity in any
positions creates incentives
Engagement Regulation way to influence voting rights, responsibilities.
for political mobilization along
5. Leadership Regulation The government reserves certain executive, these lines. Reifies the notion
judicial, or legislative positions based on of “ethnic” leaders or an
ethnic quotas or preference policies. (Note ethnic vote.
distinctions between affirmative preferences
and restrictions.)
Space and Government policies that
6. Spatial Separation of The government legalizes any separation of
Personal restrict personal movement or
People people by ethnic group in terms of residential
Interaction interactions in any way based
areas or use of public facilities. (Note distinct-
on ethnic categories are likely
ions between reservations, ethno-federalism,
to reflect and to create bases
forced segregration, and denationalization.)
for mobilization along ethnic
7. Marriage Law The government makes any legal prohibitions lines, and to reduce inter-ethnic
on marriage across ethnic lines. contact.
Opportunties 8. Employment Regulation Official sanction of use of ethnic identity for By recognizing difference,
for Personal hiring decisions. (Note distinctions between makes this “real” and creates
Advancement affirmative preferences and restrictions.) incentives to mobilize both on
part of the preferred and ex-
9. Education Regulation Official sanction of ethnic identity for select-
cluded groups; necessitates
ion/admissions. (Note distinctions between
criteria for membership.
affirmative preferences and restrictions.)

ethnicity but also mold ethnic identities (Bailey and Telles 2006:
3). Of course, decisions about census rules, like any other
decisions about institutional design, may also be the product
of particular mobilizations, but once constructed, they are likely
to endure. A recent, cross-sectional survey of census ques-
tionnaires indicates that a large proportion of countries now
categorize their populations by ethnic group (Morning 2008).2
While many studies are concerned with the size of the
groups counted by the national census, our primary concern
is with the prior question of whether or not the census actually
makes distinctions according to ethnic categories. The choice
on the part of state leaders to not enumerate ethnicity reflects
either a commonsense that such information is not important
or a deliberate attempt to mitigate the possibilities for ethnic
mobilization and conflict based on the process of such count-
ing or the results of the census. Country cases where ethnicity
is not a central or salient dimension of political life, or where
state leaders would like to eliminate ethnicity as a basis for
mobilization, are less likely to use the census to count indi-
viduals in terms of their ethnicity.
Virtually all states also issue a national identification docu-
ment to their citizens in the form of passports and/or a national
identity card. These documents help to sort the national from
the non-national other and, particularly in the case of the pass-
port, are required for travel outside of and return to one’s own
country (Torpey 2000). Again, we are concerned with whether
the national identity card or passport also states a citizen’s
ethnic identity and consequently marks persons not merely as
nationals, but also as members of a particular group. At an
extreme, the statement of a citizen’s ethnicity, specifically
whether she was Hutu or Tutsi, on Rwandan national identity
cards is widely seen as a factor that greatly facilitated the
ethnic genocide in the country in the early 1990s (Uvin 2002).
National identity cards issued by the Sri Lankan government,
which specify whether the bearer is Tamil or Sinhalese, are
widely documented to have been used to target individuals
during violence associated with the ethnic conflict that has
plagued the country for the past couple of decades.

31
Qualitative & Multi-Method Research, Spring 2009
Politics and Authority: ful of “out” group members were able to vote, we would not
Autonomy, Voting/Civic Engagement, Leadership count this as an ethnic boundary. On the other hand, because
of the centrality of space or territory to the constitution of
State leaders aspire to project authority throughout their
many ethnic groups, where space is used as a clear connota-
territory. The very idea of the national state implies an attempt
tion of ethnic difference (e.g., “North,” “South” or “federal
to unify the population behind a cohesive identity. But again,
territory”), we sometimes use such labels as indicators of eth-
when the idea of ethnic difference becomes central within so-
nic groups. For example, if representation in a national legisla-
ciety, states may make clear distinctions about the nature of
ture is allocated by federal units, and if at least one of those
political authority that depend upon individual ethnic affilia-
units is explicitly ethnic—what we describe below as ethno-
tions, from the perspectives of both ordinary citizens and lead-
federalism—we consider this a case where leadership is also
ers. We investigate the institutionalization of ethnic autonomy,
ethnically institutionalized.
political participation, and access to positions of leadership.
We hypothesize that when political authority is officially parsed Space and Personal Interaction
out in terms of ethnic groups, this can provide the basis for
We consider the extent to which states mediate the move-
enduring identification and mobilization along such lines.
ment and interactions of people in their personal lives through
A primary concern is with the degree to which the state
the legal regulation of space, residence, and personal contact.
itself projects power uniformly across all citizens, or if power is
Clearly, such restrictions imply a high salience for ethnic cat-
institutionally delegated to particular ethnic groups and asso-
egories, and provide foundations for conflict, or may be solu-
ciated leaders. In our research, we have tried to identify in-
tions to prior conflict. At certain times in history some states
stances where through the constitution or other law, autonomy
have implemented policies of ethnic separation as regards the
is afforded to ethnic leaders of at least one distinct ethnic
use of public places and residential areas. The most striking
group to regulate marriages, judicial hearings, or other per-
instances of this are racial segregation in the first part of the
sonal behaviors. When certain forms of authority are specifi-
twentieth century in the US and in apartheid South Africa. It is
cally reserved for ethnic leaders, we identify such instances as
important to note here that while there are many instances of
examples of ethnic autonomy. Because we are interested in the
ethnic groups facing effective restrictions as regards the utili-
ways in which state institutions create and reinforce specific
zation of public facilities and common spaces across the world,
group boundaries, the mere recognition of chieftaincy on the
we are concerned here with instances where this restriction is
part of the state would not be relevant in our classification
state-imposed. In particular, we identify cases in which certain
scheme, unless the state’s institutional recognition of chief-
facilities, such as schools, are reserved for the explicit use of
taincy itself clearly distinguished at least one particular group
members of particular ethnic groups. Other forms of de facto
from any others.
residential segregation may stem from the mechanism of socio-
Second, we classify countries in terms of any official re-
economic difference, and to the extent that such differences
strictions on voting or any form of civic engagement (organiz-
are not rigidly enforced and/or named, we hypothesize that
ing, protesting, etc.) based on ethnic identities. Prohibitions
they will play a much weaker role in the constitution of dis-
against the organization of ethnically-hateful or discrimina-
tinctly ethnic boundaries. Moreover, for the purposes of cross-
tory groups would not count unless such prohibitions applied
national comparison, the identification of de facto residential
explicitly to the behavior of certain ethnic groups but not oth-
segregation is much more difficult to measure, and thus sub-
ers.
ject to greater reliability concerns.
Third, we consider any distinctions made in terms of gain-
We identify four gradations of ethnic separation of space,
ing leadership positions. In order to be consistent, and reliable
and where multiple forms exist for a single ethnic cleavage in
in our classification, we are concerned with identifying only
any decade, we code for the most extreme form. First, there are
explicit—written or explicitly articulated—rules concerning the
what we call “voluntary reservations.” These are spaces that
ethnic identities of voters or office-holders in the executive or
are specifically demarcated for ethnic groups, within which
the legislature. Analogous to job and education preference
members enjoy a degree of autonomy, but members of those
policies (see below), we look for institutions that attempt to
groups may move freely inside and outside of the borders, and
balance ethnic participation, such as formula-based power-
they are fully entitled to avoid such reservations all together.
sharing quotas. We also look for restrictions that deny mem-
Second, we code for “ethno-federalism.” In such cases, there
bers of an (otherwise legally resident) ethnic group from ob-
is a specific demarcation of a regular federal unit within a larger
taining a leadership position.
federation providing a degree of ethnic autonomy. Leading
The explicit division of political power in terms of law,
indicators of the presence of such institutions are the linguis-
vote, or leadership signifies the centrality of ethnic bound-
tic-demarcation of territorial units and the use of ethnic laws
aries, and the hypothesized effects on continued identifica-
(for example, religious law) as an official rule of law for that
tion and claims-making requires little elaboration. Nonethe-
sub-national unit. Nonetheless, country cases are coded as
less, we should reiterate that we are centrally concerned with
such only when group members are still legally free to move
explicitly ethnic boundaries. For example, if the use of poll
between federal units, and need not reside in the one associ-
taxes effectively eliminated certain ethnic groups from voting,
ated with their own ethnic group. Third, we code for actual
but according to the rules, certain members of the “in” group
cases of “forced separation,” which would include cases where
were also incidentally disenfranchised and/or if a small hand-
32
 Qualitative & Multi-Method Research, Spring 2009
there are any laws limiting the use of certain public or residen- Multiple Cleavages
tial spaces for members of some ethnic groups and not others.
A central challenge for students of ethnic politics is to
Finally, we code for “denationalization,” in cases in which mem-
determine which groups or identities should be considered
bers of certain ethnic groups are not allowed to enjoy regular
relevant or salient (e.g., Posner 2005). At times, a seemingly
citizenship identification, even though they are not nationals
intractable dilemma is that scholars have failed to identify ap-
of other countries.
propriate rules for adjudicating among the overlap of identities
As a separate form of ethnic institutionalization, we in-
that any individual may hold (Chandra 2006; Chandra and
vestigate whether the state prohibits marriage or sexual con-
Wilkinson 2008). Why, for example, should one identify “Afri-
tact across ethnic lines.
can-Americans” and not “Scandinavian-Americans” as an eth-
Opportunities for Personal Advancement: nic group? Our approach is not plagued by such problems
Employment and Education because we simultaneously investigate the multiple cleavages
that might be recognized by the state, and we are not forced to
Ethnic divisions of labor (Hechter 1978) and/or divisions
make arbitrary decisions about salience or potential overlap.
of educational opportunities are often particularly important
We allow for the possibility of six different sets of ethnic cat-
bases for social and political mobilization, which at the very
egories (language, religion, caste, indigenous, race, ethnic/
least reinforce widespread perceptions of ethnic difference.
other) and search for evidence that these categories have been
We are concerned with instances in which these divisions are
used in all of the institutions described above. While religious
legally mediated and/or mandated by the state. For both work
and linguistic categories are largely self-explanatory, we clas-
and education, states may attempt to reinforce or to redress
sify race categories as those ethnic categories explicitly de-
ethnic inequalities, but in either case, the use of ethnic distinc-
scribed in terms of physical characteristics (e.g., color) and/or
tions in regulating these sets of opportunities reflects and
referred to as “race” by given state institutions; caste catego-
creates bases for ethnic political competition. Again, it is tempt-
ries as those linked to a codified caste system recognized by
ing to try to differentiate such approaches in terms of their
religious scripture, or those referred to as “caste” by state
fundamentally just or unjust qualities, but our intent is simply
institutions; indigenous categories as those groups referred
to expose where there may be bases for mobilization, and so
to as “indigenous,” “original inhabitant,” or “natives” by in-
we avoid such assessments. We simply attempt to identify
stitutions, except when the group(s) are also commonly linked
preference and quota policies that make explicit reference to
to one of the other categories already described (for example,
members of ethnic groups. Whatever the cause or rationale of
“natives” are considered a race group in South Africa); and
their creation, for our purposes, the very existence of reserva-
ethnic/other is a residual category used when state institu-
tions for individuals based on their ethnic identity constitutes
tions refer to specific ethnic or “tribal” groups that could not
an important source of institutionalization.
be classified in one of the above-mentioned categories (for
While helpful, this approach does not completely elimi-
instance, an ethnic group that is not distinguished by use of a
nate all sources of ambiguity. In many countries and time peri-
single language).
ods, preference policies are drawn up in somewhat coded lan-
guage. For example, the Nigerian constitution often refers to Data for Sixteen Countries
mandates for government employment to reflect the “federal
Because data gathering on the history of state institu-
character” of the country. When preference policies are articu-
tions is extremely labor-intensive, we have initially gathered
lated in a manner that is consistent with the institutionalization
observations for 16 country cases (not for dozens or more
of the spatial separation of people, we code these as being
than 100 as has become typical for studies of ethnic fractional-
instances of ethnic preferences.

Table 2: Case Selection for Initial Data Collection and Analysis

ELF Africa Latin America/ North America/ Asia
Caribbean Europe
Above .7 South Africa (.88) Canada (.77) India (.88)
Burkina Faso (.71) Philippines (.86)
Nigeria (.86)
.5–.69 Brazil (.59) Thailand (.63)
Pakistan (.54)
.30–.49 Botswana (.4) Costa Rica (.46) France (.32) Sri Lanka (.422)
United Kingdom (.39)
0–.29 Lesotho (.218)
Rwanda (.26)

(1985 ELF data reported from Roeder 2001.)

33
Qualitative & Multi-Method Research, Spring 2009
ization) spanning three developing regions, and several of the
high-income countries. This initial (at times painstaking) re-
search was meant to provide the foundation for the develop-
ment of strategies and classification rules which could then be
applied to subsequent research for a broader set of countries,
and potentially over a longer period of time. We make no claims
to the representativeness of this sample for the larger universe
of all countries. Rather, we first selected several of the coun-
tries based on our own familiarity with those cases,3 and then
we expanded the sample to a size we agreed would be suffi-
ciently large to identify a wide range of variation in institution-
alized ethnicity. In expanding the sample, we tried to identify
countries that we had reason to believe varied widely in terms
of ethnic politics and across a sufficiently wide geographic
scope, and across a wide range of ethnic diversity as measured
by ELF scores (Table 2), that we could put significant stress on
our analytic constructs and associated classification rules.
We used “country-decades” as the unit of analysis, though
we have also attempted to specify the exact years in which
particular laws or policies have been enacted or repealed. We
have begun the collection of data at 1900 or the decade during
which independence was claimed, whichever came later for the
country in question. If an institution existed at any time during
the decade, we count the institution as being present during
that decade. Our dataset includes 122 country-decade obser-
vations of each institution, for each of the six cleavages.
In order to learn if institutions codified ethnicity, we at-
tempted to obtain actual primary documents wherever pos-
sible, and as a second step, we tried to attain authoritative
secondary sources based on primary analyses of relevant
documents and data and/or by contacting foreign nationals,
diplomatic representatives, and scholarly authorities.
For each country-decade, we investigated the (non)-ex-
istence of all of the aforementioned institutions. For certain
types of institutions, it is far easier to be certain that an insti-
tution exists than to know that it doesn’t; for example, in the
case of “ethnic autonomy,” if we found specific provisions
allowing for such autonomy, we knew for certain that this ex-
isted and no further research was required, but we could only
be “fairly confident” that a particular institution did not exist
if, after significant searching, we could find no evidence.
Index Creation
After gathering data on the respective institutions, we
constructed what we term an Institutionalized Ethnicity Index
(IEI). The core assumption undergirding construction of the
index is that the more institutions that codify a particular eth-
nic category, the deeper the level of institutionalization, and
likely the salience of a particular category, with greater poten-
tial for conflict. We began by creating an index for each of the

Figure 1: Institutionalized Ethnicity Index Scores in 16 Countries (1900–2006)

Botswana Brazil Burkina Faso Canada

10
5
0

Costa Rica France India Lesotho

10
5
0

Nigeria Pakistan Philippines Rwanda

10
5
0

South Africa Sri Lanka Thailand United Kingdom

10
5
0

19 00 1 920 1 94 0 196 0 19 80 20 00 19 00 19 20 19 40 1 960 1 98 0 200 0 190 0 19 20 19 40 19 60 19 80 2 000 1 900 1 92 0 194 0 19 60 19 80 20 00

iei_race iei_caste
iei_lang iei_relig
iei_indig iei_other

34

six ethnic categories, allowing for a maximum 9-point score
counting 1 point for the presence of each of the first 9 institu-
tions identified in Table 1. As indicated above and in the table,
we also noted certain instances of the degree or nature of
institutionalization, but for the basic summary scores, we do
not take these into account.
In Figure 1, we depict the IEI cleavage scores for each of
our 16 countries for the period 1900–2000, but starting in the
decade of independence for those countries not already inde-
pendent in 1900. It is possible to identify four overall patterns:
A few countries have generally very high IEI scores, with at
least one cleavage reaching scores of 5 or above: South Af-
rica, India, Nigeria, Pakistan, and Canada. At the other extreme,
five countries have consistently very low IEI scores, never
scoring higher than a 1 for more than a single decade: Botswana,
Burkina Faso, Costa Rica, Lesotho, and France. In between
these extremes are Sri Lanka, Rwanda, Brazil, Philippines, Thai-
land, and the United Kingdom. Within this group, we observe
increased institutionalization of ethnic categories during the
course of the 20th century in all of the countries except Rwanda,
where the state has attempted to de-institutionalize following
the 1994 genocide. As compared with ethnic demographic
scores (such as the ELF index reported in Table 2), we believe
these data provide a much more meaningful and historically
nuanced portrait of ethnic politics. This approach does a nice
job of summarizing in a quantitative format the types of insti-
tutions and patterns of institutional change that were described
qualitatively in a number of key studies, and will provide a
solid foundation for more rigorous analyses of the causes and
consequences of those institutions.
Conclusion
This is not the place to discuss the broader substantive
or theoretical implications of our initial research, or the more
nuanced ways in which we hope to exploit both the qualitative
and quantitative data described above. Rather, for the pur-
poses of this symposium, what we believe to be most signifi-
cant is our detailing of an approach that has tried to build
upon the insights of innovative qualitative-historical research
in a manner that will allow us, and other scholars, to carry out
more disciplined and wide-ranging comparative analyses of
critical questions. The process of developing a quantitative
index forced us to clarify the essential institutional features
that drive ethnic cognitions and behaviors, which in turn has
set us in search of historical knowledge and texts. Following
recent contributions advocating such mixed method ap-
proaches (e.g., Brady and Collier 2004), we believe the strat-
egy detailed here is a promising one for tackling the thorny
conceptual and measurement problems implied by the phe-
nomenon of ethnic politics.
Notes
1
For example, we are investigating the relationship between insti-
tutionalized ethnicity and the outbreak of ethnic civil wars, using
both quantitative analyses and country histories. We will also con-
sider the impact on patterns of economic growth, and social policy.
2
Because we seek to identify particular cleavages and to code
Qualitative & Multi-Method Research, Spring 2009
censuses in earlier decades, we could not simply use Morning’s data
for our purposes.
3
In particular, Brazil, India, and South Africa.
References
Anderson, Benedict. 1996 (1983). Imagined Communities: Reflec-
tions on the Origin and Spread of Nationalism. London: Verso.
Bailey, Stanley R. and Edward E. Telles. 2006. “Multiracial Versus
Collective Black Categories: Examining Census Classification De-
bates in Brazil.” Ethnicities 6:1, 74–101.
Barth, Frederik. 1969. Ethnic Groups and Boundaries. Boston: Little,
Brown.
Brady, Henry E. and David Collier. 2004. Rethinking Social Inquiry:
Diverse Tools, Shared Standards. Lanham, MD: Rowman &
Littlefield.
Chandra, Kanchan. 2001. “Cumulative Findings in the Study of Eth-
nic Politics.” CP Newsletter 12:1, 7–11.
Chandra, Kanchan. 2006. “What is Ethnicity and Does It Matter?”
Annual Review of Political Science 9, 397–424.
Chandra, Kanchan and Steven Wilkinson. 2008. “Measuring the Ef-
fect of ‘Ethnicity.’” Comparative Political Studies 41:4–5, 515–
563.
Easterly, William and Ross Levine. 1997. “Africa’s Growth Tragedy:
Policies and Ethnic Divisions.” Quarterly Journal of Economics
112:4, 1203–1250.
Fearon, James and David Laitin. 2003. “Ethnicity, Insurgency, and
Civil War.” American Political Science Review 97:1, 75–90.
Gellner, Ernest. 1983. Nations and Nationalism. Ithaca: Cornell Uni-
versity Press.
Hechter, Michael. 1978. “Group Formation and the Cultural Divi-
sion of Labor.” American Journal of Sociology 84:2, 293–318.
Kertzer, David I. and Dominique Arel, eds. 2002. Census and Iden-
tity: The Politics of Race, Ethnicity, and Language in National Cen-
suses. Cambridge, UK: Cambridge University Press.
Laitin, David. 1986. Hegemony and Culture: Politics and Religious
Change Among the Yoruba. Chicago: University of Chicago Press.
Marx, Anthony. 1998. Making Race and Nation. Cambridge: Cam-
bridge University Press.
Morning, Ann. 2008. “Ethnic Classification in Global Perspective: A
Cross-National Survey of the 2000 Census Round.” Population
Research and Policy Review 27:2, 239–272.
Nobles, Melissa. 2000. Shades of Citizenship: Race and the Census in
Modern Politics. Stanford, CA: Stanford University Press.
Posner, Daniel N. 2005. Institutions and Ethnic Politics in Africa.
New York: Cambridge University Press.
Roeder, Philip G. 2001. “Ethnolinguistic Fractionalization (ELF) In-
dices, 1961 and 1985.” Accessed November 2004 from
http://weber.ucsd.edu/~proeder/elf.htm.
Torpey, John C. 2000. The Invention of the Passport : Surveillance,
Citizenship, and the State. Cambridge, United Kingdom; New York:
Cambridge University Press.
Uvin, Peter. 2002. “On Counting, Categorizing, and Violence in Burundi
and Rwanda.” In Census and Identity: The Politics of Race, Ethnicity,
and Language in National Censuses, ed. David I. Kertzer and
Dominique Arel. (Cambridge, UK: Cambridge University Press),
148–175.
35
Qualitative & Multi-Method Research, Spring 2009
Design Measures of Ethnic Identity: The
Problems of Overlap and Incompleteness
Kanchan Chandra
New York University
kanchan.chandra@gmail.com
A great deal of new work in the field of ethnic politics has
focused on improving the quality of the data we put into our
measures—and critiquing the quality of the data that went
into measures such as the Index of Ethnolinguistic Fractional-
ization (the ELF index) (Chandra 2009, Chandra and Wilkinson
2008, Cederman and Girardin 2007, Posner 2004).1
In this article, I shift the focus to improving the quality of
the assumptions that drive our measures. The ELF index and
many of the new measures which have followed it are based on
the assumptions that the ethnic categories that describe a popu-
lation are mutually exclusive. But the membership of one eth-
nic category, whether nominal or activated, often overlaps with
the membership of another. I refer to this as the problem of
“overlap.” The ELF index and some additional measures also
assume that the ethnic categories that describe a population
are exhaustive. But it is often the case that only a part of popu-
lation can be categorized in ethnic terms. I refer to this as the
problem of “incompleteness.” Quite apart from improving the
quality of the data that they describe, improving the quality of
our measures requires us to improve the quality of our as-
sumptions to take these problems into account.
I then introduce one new measure of the activation of
ethnic identities by political parties—EVOTE—that does not
employ the assumptions of exclusiveness and exhaustiveness.
EVOTE is drawn from a time-series cross-national dataset that
I am currently in the process of constructing (referred to here-
after as CDEI (Constructivist Dataset on Ethnicity and Institu-
tions). CDEI collects data on ethnic categories activated in
competitive politics by political parties and the institutional
context in which such politicization takes place. The full do-
main of CDEI is all countries that held at least one lower-house
legislative election in any of three decades: 1976–1986, 1986–
1996, and 1996–2006. The data and examples in this article
however, are taken from the 1986–1996 cross-section. They
describe a hundred countries which held a party-based elec-
tion for the lower-house of the legislature between 1986 and
1996.
EVOTE is only one of many measures that are or could be
designed without making the restrictive assumptions of exclu-
siveness and exhaustiveness. Other such measures we are
developing in CDEI include measures of the effective number
of ethnic parties, of the degree of dispersion of ethnic group
vote across ethnic parties that represent it, of the degree of
inclusion of ethnic parties in government, and of the dimen-
sions to which ethnic categories activated by political parties
belong. These were introduced and discussed at a workshop,
“Measuring Ethnicity,” held at New York University in Octo-
ber 2008. Other scholars are also developing measures that do
36
not, as far as I can tell, assume exhaustiveness and exclusive-
ness. These include the Index of Institutionalized Ethnicity
(IEI), proposed by Lieberman and Singh in this symposium
(Lieberman and Singh 2008), ECI, proposed by Steven Wilkinson
(Chandra and Wilkinson 2008), among others.
What is an Ethnic “Group”?
The Oxford English Dictionary associates the word
“group” with two meanings: “A number of people or things
regarded as forming a unity or whole on the grounds of some
mutual or common relation or purpose,” or “classed together
because of a degree of similarity.” The second meaning sug-
gests that those who belong to a group simply share member-
ship in a category—a descriptive label that describes and dis-
tinguishes them from others regardless of their own feelings
on the matter. The first meaning suggests that those who share
group membership not only share a descriptive label but also
think of themselves as a collective with a shared conscious-
ness.
The identities that comparative political sciences classify
as ethnic are often no more than descriptive categories—but
our theories, often unjustifiably, associate members of these
categories with a shared collective consciousness (Brubaker
2004). In my own work, and in the course of this article, I treat
an ethnic identity purely as a descriptive category and try to
avoid using the ambiguous term “group” altogether. But when
referring to the usage of other comparative political scientists
who do use the word “group,” I also follow suit, trying either
to clarify what I think it means in the work from which it is
drawn or consider all possible interpretations of what it could
mean.
By an “ethnic” category, I mean a category in which mem-
bership depends on a subset of descent-based attributes. All
categories based on descent-based attributes, according to
this definition, are not ethnic identity categories. And some
non-ethnic identities also require descent-based rules for mem-
bership. But all ethnic identities require some descent-based
attributes for membership. This definition captures the classi-
fication of ethnic identities in comparative politics to a greater
degree than the alternatives (for elaboration, see Chandra 2006,
2007).
Nominal ethnic identities are those ethnic identity cat-
egories in which an individual’s descent-based attributes make
her eligible for membership. Activated ethnic identities are
those ethnic categories in which she actually professes mem-
bership or to which she is assigned membership by others. All
individuals have a repertoire of nominal ethnic identities from
which one or more may be activated.2
The ethnic “structure” of a population consists of all the
repertoires of nominal ethnic identity categories of all indi-
viduals in that population—and the attributes from which these
nominal categories are generated (for elaboration, see Chandra
2009 and Chandra and Wilkinson 2008). Suppose, in an ex-
ample to which I will return throughout this article, individuals
in a population are characterized by two types of skin colour:
black and white. And suppose that they are characterized by
two places of origin: foreign and native (Chandra and Boulet

2006). These attributes can generate a very large number of
nominal categories, including “Blacks” (consisting of those
individuals with the attribute-repertoires black and foreign or
black and native), “Whites” (consisting of those individuals
with the attribute-repertoires white and foreign or white and
native), “White Natives” (consisting only of those individuals
who are both white skinned and native), “Foreigners” (con-
sisting of those individuals who are either white skinned and
foreign or black skinned and foreign), “Black Natives” (con-
sisting only of those individuals who are both black skinned
and native), and so on. All these nominal categories belong to
this population’s ethnic “structure” whether or not anyone
actually identifies with them.
By ethnic “practice,” I mean those categories actually ac-
tivated by a country’s population in a specific context. The
context can vary: the categories activated in party politics can
be different from categories activated in voting behaviour and
both can be different from categories activated in private life.
Suppose, for instance, that in the population above, individu-
als who have the attributes of black skin and a foreign place of
origin) describe themselves as “Black” during elections—al-
though they also qualify for membership in the nominal cat-
egories of “Foreigner” or “Black Foreigner,” these are not the
ones in which they declare membership. We would then term
the category “Black” as the activated category in this context.
The Assumptions of Mutual Exclusiveness and
Exhaustiveness in the ELF Index and Other Measures
The ELF Index is calculated according to the formula 1–
si2, where si is the proportion of the ith activated ethnic
category, i={1, 2, …..n}. This formula requires the ethnic cat-
egories to be mutually exclusive (i.e., if you are in ethnic cat-
egory 1, you are not in ethnic categories 2–n) and exhaustive
(every member of the population is in some ethnic category).
Given mutual exclusiveness and exhaustiveness, this index
measures the probability that two randomly chosen individu-
als from a country’s population belong to different groups.
Thus, a society with two groups, a majority of 80% and a
minority of 20%, would have an ELF score of 1–(.64+.04) = .32.
A society with several small groups of 25% each would have
a higher ELF score of 1–(.0625+.0625+.0625+.0625) = .75. In
the absence of mutual exclusiveness and exhaustiveness, this
measure would be meaningless: it would no longer measure
the probability that two randomly chosen individuals from a
country’s population belong to different groups.
The Problem of Overlap
The “attributes” from which a country’s nominal or acti-
vated ethnic categories are generated can indeed be arrayed
on one or more mutually exclusive attribute dimensions. Thus,
if we define as “foreign” as anyone whose parents were born
outside the boundaries of a country, and “native” as all those
whose parents were not born outside the country, we have a
mutually exclusive attribute dimension of place of origin where
everyone is either foreign or native. If there are people who
don’t quite fit, we can simply place them in a residual value.”3
But the ELF index and related measures purport to de-
Qualitative & Multi-Method Research, Spring 2009
scribe the ethnic identity categories that characterize a popu-
lation—not the attributes from which they are generated.
These categories are often not mutually exclusive. To illus-
trate, consider first the nominal categories in our hypothetical
example above. Some of these categories such as “Black” and
“White,” indeed have mutually exclusive memberships, be-
cause the attributes that qualify individuals for membership in
each—the attribute of “black” skin for the category Black and
the attribute of “white” skin for the category White—are mu-
tually exclusive by definition. But many others are not. Con-
sider the categories White (consisting of those individuals
with the attribute-repertoires white and foreign or white and
native) and “White Natives” (consisting only of those indi-
viduals who are both white skinned and native). These cat-
egories have overlapping memberships: given their attributes,
those who are White can also be described as “White and
Native.”
The real world is full of such examples. The United States
is perhaps the closest example to the hypothetical case above.
Nominal ethnic identity categories in the US include “White,”
“Black,” “WASP,” “Immigrant,” “Irish-American,” “Latino,”
“Catholic,” “West Indian,” and so on. The categories “White”
and “WASP” overlap, as do the categories “White,” “Irish-
American,” and “Catholic.” The categories “Black,” “Catho-
lic,” and “West Indian” overlap too. So do the categories
“Immigrant,” “Black,” “Irish-American” and “Latino.” We could
reproduce examples of such overlap when surveying the eth-
nic structure of most countries.
Suppose the term “group” refers to the activated catego-
ries embedded in a population’s ethnic “practice.” There is no
logical reason to expect individuals to activate only mutually
exclusive categories—and we do not so far have a theory that
makes such a prediction. In our hypothetical example, sup-
pose that individuals activate the categories “Foreign,” “Black,”
and “White Native.” These categories have overlapping mem-
berships: some of those who have the attributes for member-
ship in the category “Foreign” also qualify for membership in
the category “Black.”
Indeed, at least judging from CDEI data on ethnic practice
in party politics, overlapping categories routinely show up in
party politics. Very occasionally, we do find cases in which
activated categories belong to a category-dimension. In
Guyana, for instance, the two activated categories in our data—
Afro-Guyanese and East Indian—could be said to be mutually
exclusive and belong to a single dimension based on region of
origin. But these are exceptions that prove the rule. In most of
the 100 countries that we study, the ethnic categories acti-
vated do not belong to any commonsensical family. CDEI’s
count of politically activated ethnic categories in India, for
instance, produces the following categories: Hindu, Muslim,
Sikh, OBCs, Scheduled Castes, Jharkhandis, Assamese, and
Tamils. The categories Hindu, Muslim, and Sikh belong to the
attribute-dimension of religion, the category OBC and Sched-
uled Caste belong to the attribute-dimension of caste, the cat-
egories Assamese and Tamils belong to the coinciding dimen-
sions of region and language, and the category Jharkhandi
belongs to the dimension of subregion. Many of them overlap
37
Qualitative & Multi-Method Research, Spring 2009
with each other. To cite just a few examples, the membership of
the categories Hindu, Muslim, Scheduled Caste, Assamese
Tamil, and Jharkhandi overlaps. So does the membership of
the categories Muslim, OBC, Tamil, and Assamese. In Belgium,
the activated categories are Flemish Speakers, French Speak-
ers, Walloons, Brussels, French-speaking Brussels, Native-
Flemish-Belgian, Native-French-Belgian, and German. The cat-
egory Flemish Speakers overlaps with the category Native Bel-
gians, as does the category Walloon.
The Problem of Incompleteness
Is it reasonable to believe that counts of ethnic “groups”
represent an exhaustive description of individuals in a popula-
tion? If by the term ethnic “group” we mean all the nominal
ethnic identities in a country’s ethnic structure in which indi-
viduals are eligible for membership, then the answer is yes. All
individuals have nominal ethnic identities, whether or not they
actually activate them. Thus a population can be described
exhaustively by a count of nominal ethnic identity categories,
although, for the reasons articulated above, this exhaustive
count may consist of overlapping categories.
But if by ethnic “group” we mean the categories activated
in ethnic “practice,” then the answer is no. There is no reason
to expect that all individuals in a population should activate
ethnic identity categories. To illustrate, consider the study of
ethnicity in a Translyvanian town conducted by Brubaker et
al. (2006). This study found that in many contexts, those who
were nominally members of the Romanian ethnic category did
not activate their Romanian ethnic identity, while those who
were of Hungarian ethnicity did. Indeed, in many countries,
including the US, UK, Myanmar, and India, majority groups
often do not activate ethnic identities—the word “ethnic” is
typically reserved for “minorities.”
Indeed, the data in CDEI shows that there are only a few
very polarized countries at particular points in time, such as
Yugoslavia in 1992, where almost the entire population lines
up behind parties activating an ethnic identity—but even in
such countries, the ethnic identification may not be complete.
CDEI shows that 86% of the population in Yugoslavia voted
for ethnic parties in 1992, leaving a minority of voters who
voted for other types of parties. Even if we make the strong
assumption that voting for an ethnic party means activating
an ethnic identity category, this suggests that a significant
proportion of the population did not activate an ethnic iden-
tity category at all. In most countries in the dataset, voters
voted for parties that activated a wide range of categories,
non-ethnic as well as ethnic. In the recent US presidential elec-
tions, for instance, some voters appear to have activated class
identities (e.g., middle-class), others their party identities (Re-
publican or Democrat), others identities based on age (e.g.,
pensioners or young people) and still others their racial identi-
ties (e.g., Black).
Implications of the Problems of
Overlap and Incompleteness for our Measures
The preceding sections argue that if by “ethnic groups”
we mean all the nominal ethnic groups embedded in a country’s
38
ethnic structure, then they can in many instances not be taken
to be mutually exclusive. They can in principle be exhaustive.
But because this exhaustive set of nominal categories includes
those with overlapping memberships, it cannot be accurately
summarized through the ELF index and other comparable mea-
sures.
The categories activated in ethnic practice, by contrast,
are often neither mutually exclusive nor exhaustive. As mea-
sures of either ethnic structure or ethnic practice broadly de-
fined, then, measures that employ these assumptions are mean-
ingless. They distort rather than simplify the empirical world
that they aim to describe and generate uninterpretable re-
sults.
Does this mean that we should not employ the assump-
tions of exclusiveness and exhaustiveness in our measures?
No—but we should recognize that we can only employ these
assumptions to describe specific concepts in specific condi-
tions and interpret our results accordingly. We could, for in-
stance, indeed employ the assumptions of exclusiveness and
exhaustiveness to summarize any single attribute-dimension.
For example, if we were describing India’s ethnic structure, we
could probably come up with mutually exclusive lists of at-
tributes on the attribute-dimensions of tribe, religion, lan-
guage, race, nationality, caste, and region. There may well not
be a single “correct” level of aggregation on these dimen-
sions, and some attributes will belong to multiple dimensions.
But I can imagine some systematic rules that allow us to con-
struct these attribute-dimensions.
This list of attribute-dimensions is likely to include more
than one dimension per country. In the example I just gave,
there are two attribute-dimensions: skin color and place of
origin. In the real world, many countries have multiple attribute
dimensions. In the case of India, the dimensions I can think of
are at least seven. In the US, they are at least six, including
race, religion, language, region, tribe, and nationality. In Ma-
laysia, similarly, there are at least five, including race, language,
region, religion, and tribe. We will need, therefore, a statistical
measure of ethnic “structure” which, unlike the ELF index, is
designed to capture this multi-dimensionality. Even if we were
able to design such measures, furthermore, we would need to
modify our interpretations of statistical results. Summarizing
the repertoire of attribute-dimensions in a population’s ethnic
structure is not the same thing as summarizing the repertoire of
ethnic identity categories in a country’s structure.
Alternatively, we might want to consider attribute-dimen-
sions one by one, considering, as Alesina et al. (2003) and
Fearon and Laitin (2003) do, the effect of the dimensions of
language or religion or other such dimensions separately. A
concentration index like the ELF index could be used in this
case, since it eliminates the problem of overlap. But here, too,
we would need to modify our interpretations carefully. Results
generated from a consideration of an individual attribute-di-
mension embedded in a country’s ethnic structure should be
interpreted as narrow results about specific attribute-dimen-
sions—not as general results about “ethnicity.” Further, they
should be interpreted as results about nominal rather than
activated dimensions, since we have no reason to believe that

activated identities fall on a single dimension. Finally, they
4
should be interpreted as results that apply not to ethnic iden-
tity categories themselves, but to the raw materials from which
those categories are generated.
A New Measure: EVOTE
The EVOTE measure is one of those generated from the
CDEI. It measures the total percentage of the vote share ob-
tained by ethnic parties in a given country, taken together, in
the lower-house legislative election closest to 1996 in the 1986–
1996 decade. EVOTE is obtained by aggregating the votes
obtained by individual ethnic parties in a given country (EVOTE
= Vote for Ethnic Party 1 + Vote for Ethnic Party 2 + Vote for
Ethnic Party 3 ….).
This measure is constructed as follows: for each country
included in the dataset, we first obtain a list of parties and vote
shares at as disaggregated a level for the relevant legislative
election. We then classify each political party in each country
for which vote shares are available according to whether it
activates a target category based on ethnicity in several ways:
does it activate ethnicity in its name, in its platform, in its
support base (defined two ways), or in its leadership.
The EVOTE measure discussed here refers entirely to the
votes won by parties that are classified as ethnic by platform.
But in principle, we could construct the same measure by ag-
gregating the votes of parties that are ethnic by name, by
support base (differently defined), and by leadership.
The classification of parties is based on a content analy-
sis of the election campaign of the party in question using four
sources: the Europa World Yearbook, the Political Hand-
book of the World, news sources from FBIS (Foreign Broad-
cast Information Service), and LexisNexis searches. For each
party, we obtain a sample of campaign materials (speeches at
election rallies, policy pronouncements, and so on) as reported
in FBIS and LexisNexis for a period up to three months before
Figure 1: Frequency Distribution of EVOTE
6040
Percent
20
0
0 20
Qualitative & Multi-Method Research, Spring 2009
the election date. These include reports from the international
media and translations of local news reports from newspa-
pers, radio and TV. These samples are primary sources that
report what parties are actually saying to voters at the time of
the election. Where the samples are too small to permit reliable
codings, we turn to local newspapers and secondary sources
as a last resort. These sources give us a sample of articles for
the election platform of each party individually.
We code each sample for each party by platform in the
following way: If we find that a political party makes an open
and exclusive appeal to some ethnic category or set of catego-
ries, and that such an appeal is central to its campaign, we
code it as an ethnic party. If we find that a political party makes
an open and inclusive appeal to all ethnic categories that de-
fine a population and makes such an appeal central to its elec-
tion campaign, we code it as a multi-ethnic party. And if we
find that a political party does not make an open or a central
appeal to an ethnic category, whether exclusive or inclusive,
we code it as a non-ethnic party.
Consider the case of India as an example of our coding
procedures. Hundreds of parties competed in India in the 1991
parliamentary elections (the elections closest to but before
1996) but most of them obtained a miniscule percentage of the
vote. We obtained disaggregated data on all parties that ob-
tained at least .01% of the vote, thus including 66 parties in
our dataset. (Note that this means that our dataset excludes
both parties which did not contest a particular election, and
parties that obtained less than .01% of the vote. As such, it
undercounts small and failed parties). We then coded each of
these 66 parties based on a content analysis of its party plat-
form. Of the 66 parties, we coded 13 parties, accounting for
51.81% of the vote, as non-ethnic, 18 parties, accounting for
38.95% of the vote, as ethnic, and were not able to find suffi-
cient articles on election platforms to code the remaining 35
parties, accounting for 10.24% of the vote (these were very
40 60 80
evotep
39
Qualitative & Multi-Method Research, Spring 2009
small parties, with a mean vote of .14%).
When a party made an explicit ethnic appeal, we also col-
lected data on the name of its target ethnic category and the
size of the population that each ethnic category constitutes.
In India, for instance, ethnic parties, taken together, explicitly
mobilized the following ethnic categories: Hindus (82%), Mus-
lims (12.12%); Sikhs (1.94%); OBCs (52%), Scheduled Castes
(16.48%), Jharkhandis (3.18%), Assamese (2.64%), and Tamils
(6.6%).
In these data, EVOTE ranges from 0 (in 44 out of 100 cases)
to a maximum of 85.63 (for Yugoslavia in 1992, with Israel in
1992 and Zimbabwe in 1995 close behind), with a mean of 12.95.
The graph below represents the frequency distribution of
EVOTE in these data.
EVOTE is a precise measure of the politicization of ethnic
identities in practice and not simply of ethnic structure. As a
measure of ethnic practice, it does not require the assumptions
that ethnic categories are mutually exclusive and exhaustive
when they are often not. To illustrate, let’s return to the cat-
egories activated in the 1991 elections in India. Of these, the
categories Hindu, Muslim, and Sikh are mutually exclusive in
relation to each other, but overlap with the categories OBCs,
Scheduled Castes, Jharkhandis, Assamese, and Tamils. The
percentage of the vote captured by ethnic parties also indi-
cates that they do not exhaust the categories activated by the
population: to the extent that the majority of the vote in these
elections went to non-ethnic parties, we must assume that a
significant proportion of the population activated identity cat-
egories other than those defined by ethnic identity.
Given that categories in practice are often neither exhaus-
tive nor mutually exclusive, putting them into the ELF index
yields a nonsensical number. By contrast, EVOTE remains mean-
ingful without these assumptions. The total proportion of the
vote won by ethnic parties is unaffected by whether the ethnic
parties in question mobilize mutually exclusive or overlapping
categories. In India, it so happens that the parties activate
overlapping categories. But the value of EVOTE would be the
same even if the parties in question activated mutually exclu-
sive categories. Similarly, EVOTE also does not impose the
requirement that the categories activated by political parties
be complete, since it is the votes won by the parties that mobi-
lize each category that are added, not the proportion of the
population made up by the categories themselves. Rather, it
allows us to observe such completeness in the data. In coun-
tries in which all individuals activate ethnic identities in their
voting behaviour, the value of EVOTE would be 100%. In coun-
tries in which only some individuals activate ethnic identities
in their voting behaviour, the value of EVOTE would be less
than 100%.
Perfection vs. Interpretability
EVOTE is a narrow measure of the activation of ethnic
identity by political parties in their explicit appeals. I expect to
use EVOTE to examine whether such explicit use of categories
by political parties is associated with breakdown and instabil-
ity. But it is not an all-purpose measure. It conveys no infor-
mation about the effect of other ways in which ethnic identity
40
categories may be politicized, either through the implicit
behaviour of political parties, or through the behaviour of vot-
ers rather than parties or through the use of ethnic identity in
non-electoral contexts. A study of the relationship between
EVOTE and breakdown and stability, thus, is informative about
the relationship between one aspect of ethnicity and these
variables, but not about all aspects. Designing an array of
such narrow concepts and measures is a more promising strat-
egy to explore the effect of ethnicity than designing a single
all-purpose measure of “ethnicity.”
Even as a narrow measure, EVOTE has plenty of biases
and limitations that need to be corrected for or at a minimum
taken into account in the interpretation of results generated
from it. To cite only two from a long list, it systematically un-
der-counts small parties, and measurement error in the coding
of ethnic parties from which it is generated is correlated with
the size of the party and the region it is from. Using EVOTE for
individual analyses will surely uncover instances of bias and
error that we have not considered so far. But the main case for
EVOTE—and other measures that abandon the assumptions
of exhaustiveness and exclusiveness—is not that it is perfect,
but that it is interpretable. That is a minimal standard, worth
defending in a field where it is too often ignored.
Notes
1
This article makes arguments developed more fully in an ongoing
book manuscript and introduced in several articles, including Chandra
(2006, 2008, 2009), Chandra and Wilkinson (2008).
2
I use this precisely defined distinction between “nominal” and
“activated” categories rather than the similar-sounding distinctions
“latent” and “salient” ethnic identities, “dormant” and “mobilized
identities, and “commonsensically real” and “politically relevant”
identities (Posner 2004, 2005). Salient or mobilized or politically
relevant identities or group memberships are often taken to mean not
only shared membership but the sharing of some content, such as
common preferences or culture or symbols. But as I use it here,
“activating” an ethnic category simply requires an individual to claim
membership in it—it does not require her to subscribe to its content,
or even necessarily to associate with other members of that category,
although she may also do that.
3
Note that there is nothing “objective” about how we define and
group values on a common dimension. We might just as easily define
as foreign those who are themselves born abroad, or those whose
ancestors in the last 100 years were born abroad, etc. But the main
point here is that we can, and usually do, construct subjective defini-
tions that group attributes into mutually exclusive families. These
attributes then become the raw materials by which nominal and acti-
vated categories are constructed.
4
For a different view, see Posner (2005), which is premised on the
view that activated categories fall into mutually exclusive “category-
sets” or “dimensions.”
References
Alesina, Alberto, Arnaud Devleeschauwer, William Easterly, Sergio
Kurlat, and Romain Wacziarg. 2003. “Fractionalization.” Journal
of Economic Growth 8:2, 155–194.
Brubaker, Rogers, Margit Feischmidt, Jon Fox, and Liana Grancea.
2006. Nationalist Politics and Everyday Ethnicity in a Transylvanian
Town. Princeton, NJ: Princeton University Press.

Brubaker, Rogers. 2004. Ethnicity Without Groups. Cambridge:
Harvard University Press.
Cederman, Lars-Erik, and Luc Girardin. 2007. “Beyond Fractional-
ization: Mapping Ethnicity onto Nationalist Insurgencies.” Ameri-
can Political Science Review 101:1, 173–186.
Chandra, Kanchan. 2006. “What is Ethnic Identity and Does it
Matter?” Annual Review of Political Science 9, 397–424.
Chandra, Kanchan. 2007. “Part I: A Conceptual Framework for Think-
ing About Ethnic Identity.” In Ethnicity, Politics and Economics
(Manuscript in progress).
Chandra, Kanchan. 2008. “Making Causal Claims About the Effect
of ‘Ethnicity.’” In Mark Lichbach and Alan Zuckerman, eds.,
Comparative Politics: Rationality, Culture, and Structure. Cam-
bridge: Cambridge University Press.
Chandra, Kanchan. 2009. “A Constructivist Dataset on Ethnicity
and Institutions.” In Rawi Abdelal, Yoshiko Herrera, Ian Johnston
and Rose McDermott, eds., Identity as a Variable. New York:
Cambridge University Press.
Chandra, Kanchan and Cilanne Boulet. 2006. “A Language for Think-
ing about Ethnic Identity Change.” In Ethnicity, Politics and Eco-
nomics. (Manuscript in progress).
Chandra, Kanchan and Steven Wilkinson 2008. “Measuring Ethnicity.”
Comparative Political Studies 41:4, 515–563.
Fearon, James D. and David D. Laitin. 2003. “Ethnicity, Insurgency
and Civil War.” American Political Science Review 97:1, 75–90.
Lieberman, Evan and Prerna Singh. 2008. “Institutionalized Ethnicity.”
Paper presented at the Annual Meeting of the American Political
Science Association.
Posner, Daniel N. 2004. “Measuring Ethnic Fractionalization in Af-
rica.” American Journal of Political Science 48:4, 849–863.
Posner, Daniel N. 2005. The Institutional Origins of Ethnic Politics in
Africa. Cambridge: Cambridge University Press.
Measuring Identity Content and
Comparing Social Identities
Yoshiko Herrera
University of Wisconsin-Madison
yherrera@wisc.edu
Rose McDermott
Brown University
Alastair Iain Johnston
Harvard University
johnston@fas.harvard.edu
Rawi Abdelal
Harvard Business School
rabdelal@hbs.edu
This symposium is on the one hand a testament to the
incredible progress that has been made in the measurement of
social identities, and in particular ethnic identities, over the last
several years; on the other hand, the articles also highlight
how much work is yet to be done.1 Our contribution to this
debate is to focus on content and comparison of social identi-
ties: we make a case for why the content, or meaning, of social
identities matters, and how measuring content will help us with
another major task, which is to be able to make comparisons
Qualitative & Multi-Method Research, Spring 2009
across types of identities, e.g., ethnicity, race, caste, religious
association, gender, class, etc. We begin with a discussion of
why content matters and how it is related to the ability to make
comparisons. We then outline the “Identity as a Variable” ana-
lytical framework for measuring the content and contestation
in social identities (Abdelal et al. 2006, 2009) and we discuss
the variety of methodological techniques that can be used to
measure identity content by briefly describing the contribu-
tions to our recently published edited volume, Measuring Iden-
tity: A Guide for Social Scientists, which illustrates several
different ways that scholars have operationalized the mea-
surement of the content of ethnic and other identities.
Why Content Matters
Measurement of ethnic identity means many things. As
Kanchan Chandra points out in this symposium, there is a
theoretically significant difference between nominal and acti-
vated ethnic identities, and the measurement techniques for
addressing each needs to be different. In addition, we can also
add the concepts of groupness and content to what might be
measured by ethnic identity, both of which are analytically
distinct from lists of nominal or activated identities.
“Groupness,” along the lines proposed by Brubaker and
Cooper (2000), describes the degree to which individuals iden-
tify with a group. At a minimum, groupness is merely a di-
chotomous variable—either groupness exists, or it does not
(e.g., the survey question, “Check this box if you are African-
American”). Or, there may be degrees of groupness—measur-
ing how much individuals identify with a given group (e.g.,
the survey question, “How strongly do you identify as Afri-
can-American?”). Whatever way one measures groupness, it
is almost certainly going to fall victim to the problem of over-
lap as discussed by Chandra; that is, the list of groups in a
question is unlikely to capture all the possible identities that a
person might have, or it is likely to treat them as mutually
exclusive. More importantly, the conception of groupness
takes the meaning of a group among its members for granted
and therefore ignores the content of the identity, i.e., the shared
behavioral expectations, goals, views of other groups, and
interpretative processes that constitute the meaning of the
group. In addition, the meaning or content of identity varies
across individuals within groups, and that too is excluded by
the concept of groupness. Hence, the concept of groupness,
by sidestepping the question of content or what the group
means to its members, ignores one of the most theoretically
and empirically significant aspects of social identity.
The content or meaning of social identities, we argue, is
the mechanism which links having an identity to specific be-
havior, to the formation of interests, and to relations with other
groups. For example, if we consider a foreign-born, black, fe-
male, Christian, New Yorker, we need to know the meaning of
all those labels to say anything about this woman’s behavior
(e.g., who she will likely vote for, where she will choose to live,
what languages she will speak, what kind of food she will eat,
what holidays she will celebrate, etc.). The content of her iden-
tities is also the key to her interests and goals, such as which
party she identifies with, whether she values education, which
41
Qualitative & Multi-Method Research, Spring 2009
charities she supports, and which political issues she will get
involved with. The identity content will tell us something about
her likely relations with other groups, such as whites, foreign-
ers, Catholics vs. Muslims, etc., and finally, content will likely
tell us something about how this person hierarchically sorts
out her multiple identities, i.e., which one is most important to
her. We believe that the framework we outline below that de-
tails types of identity content, and contestation over that con-
tent, allows for assessment of the meaning of social identities.
Comparisons of Types of Identities
A serious conceptual challenge in the identity literature
concerns the question of how to compare different types of
identities, e.g., ethnicity, race, national identities, gender, caste,
religious identities, and class.2 We know that individuals often
have more than one type of identity and sometimes multiple
identities within the same type (e.g., more than one ethnic iden-
tity), and these identities may be hierarchical (i.e., one identity
is more important than another). To acknowledge that there are
different types of identities brings up the theoretical question
of what differentiates these identities and what they have in
common, if anything. In addition, there is the empirical ques-
tion of whether a dataset should include all of these types of
identities (or a subset of types) or whether they should be
treated separately.
The literature thus far is not, in our opinion, entirely satis-
factory. Some datasets purport to measure one thing, e.g., eth-
nic categories only, but include others such as religious cat-
egories. For example, Russian census data claims to exclude
religion, but includes “Jewish” and a few other religious cat-
egories on the idea that some religious identities are actually
ethnic identities. In addition, some datasets explicitly claim to
measure more than one category, but are widely used as a proxy
for one; e.g., the Ethnolinguistic Fragmentation Index (ELF)
has both ethnicity and language in its title but is perhaps the
most widely measure of “ethnicity.” These cases highlight the
blurry conceptual distinction between religion, ethnicity, and
language. An alternative is to focus only on one type of iden-
tity, say language or ethnicity (Alesina et al. 2003). Yet when
this is done regarding ethnicity, we know it to be problematic
because of the fact that focusing only on ethnicity excludes
other politically relevant comparable identities, and that fact is
often not reflected in the interpretation of the data (as Chandra
discusses in this symposium and elsewhere in more detail; see
Chandra 2006).
Finally, some scholars broaden the term ethnicity to en-
compass other types of identity, both in theory and in practice.
Donald Horowitz writes, “Ethnicity easily embraces groups dif-
ferentiated by color, language, and religion; it covers ‘tribes,’
‘races,’ ‘nationalities,’ and ‘castes’” (Horowitz 1985: 53). On
the empirical level, in the Index of Institutionalized Ethnicity
(IEI) discussed by Lieberman and Singh in this symposium,
ethnic categories include “language, religion, caste, indigenous,
race, and ethnic/other” categories. Similarly, EVOTE, discussed
by Chandra in this symposium, includes multiple types of iden-
tities (or dimensions, in her terminology). We agree that it is
intuitively sensible to broaden the concept of ethnicity in par-
42
ticular data sets, because otherwise theoretically important
categories would be left out.
Nevertheless, while broadening the term ethnicity to mean
multiple types of identity is one solution, we suggest an alter-
native, namely to disaggregate types of identity content so
that comparisons across identity types can be made based on
shared types of content. For example, social norms regarding
language and dietary restrictions can be compared among eth-
nic and religious groups; worldviews or cognitive schemas
can be compared across gender, class, and racial groups;
shared goals could be compared across religious or national
identity groups, and so on. Moreover, by concentrating on
the comparison of types of content, such as relational com-
parisons, we can squarely address concepts like status hierar-
chies, which might include caste, religion, and ethnicity.
Brubaker et al. (2004) attempted to compare and differen-
tiate race, ethnicity, and national identities, arguing that cogni-
tion is a way to address all three identity types. While we do
not disagree that cognition or cognitive content of identities
may be a way to compare identity types, we argue that cogni-
tion is not the only comparable type of content and it is also
not just race, ethnicity, and national identity which demand
comparison and differentiation from each other. As the ex-
amples above suggest, scholarly work on all social identities
might be better served by an analytic framework that allows
for comparison and differentiation among identity types, and
we argue that focusing on types of content is the key to such
a framework.
An Analytic Framework for Measuring the
Content and Contestation of Identities
Toward the development of an analytic framework that
allows for comparison and differentiation among the many
types of identities, we offer a definition of a collective identity
as a social category that varies along two dimensions: content
and contestation. Content describes the meaning of a collec-
tive identity. Moreover, the content of social identities may
take the form of four nonmutually exclusive types: constitu-
tive norms, social purposes, relational comparisons with other
social categories, and cognitive models. Contestation refers
to the degree of agreement within a group over the content of
the shared category. Our conceptualization thus enables col-
lective identities to be compared according to the agreement
and disagreement about their meanings by the members of the
group.
Constitutive Norms
In brief, norms are behavioral expectations for a given
identity group (Jepperson et al. 1996). The normative content
of a collective identity specifies its constitutive rules—the
practices that define that identity and lead other actors to
recognize it. For example, for an ethnic group this kind of con-
tent could include “rules” on language use, religion, dress
codes, etc. Note that types of identities can be embedded in
norms, e.g., Armenians are also Christians; to convert to Islam
would make others in the group seriously question one’s “Ar-
menian-ness.” Also, different types of identities might have

different norms, but there could be some overlap: Religious
identities, for example, might have norms that include dietary
restrictions, birth, or marriage ceremonies, but they might also
include dress codes or language use similar to ethnic group
norms. All of the rules that determine group membership and
putative attributes of the group can be thought of in terms of
constitutive norms, which can be informal or formal so long as
they set collective behavioral expectations for members of the
group. Moreover, these constitutive identity group norms likely
derive from a broader set of social norms that emanate from
multiple centers of authority. Hence constitutive norms also
link particular identity groups to larger historical and social
contexts.
Social Purposes
The content of a collective identity may be purposive, in
the sense that the group attaches specific goals to its identity.
This purposive content is analytically similar to the common-
sense notion that what groups want depends on who they
think they are. Thus, identities can lead actors to work towards
particular group purposes or goals. For example, most national
identity groups share the goal of supporting their nation-state,
or in the case of nationalist movements, establishing a nation
state. Similarly, a religious identity might include shared pur-
poses, such as working to increase converts, reduce poverty,
or end the death penalty, etc. Whereas the normative content
of an identity refers to practices that lead to individual obliga-
tion and social recognition, the purposive content of an iden-
tity helps to define group interests, goals, or preferences. Both
the normative and purposive content of an identity may im-
pose obligations on members, but in distinctive ways: consti-
tutive norms impose an obligation to engage in practices that
define the group, whereas social purposes create obligations
to engage in practices that make the group’s achievement of a
set of goals more likely.
Relational Comparisons
The content of a collective identity includes relational
comparisons when the identity is defined in terms of refer-
ences to other collective identities. In other words, an identity
may be defined by what it is not—that is, by some other iden-
tities. Relational content of collective identities can include,
for example, the extent to which one social identity excludes
the holding of another (exclusivity); the relative status of an
identity compared to others; and the existence or level of hos-
tility presented by other identities. Gender and class identities
present obvious illustrations of relational content; even the
category of “transgender” references sex-based differences
between men and women, and it’s hard to imagine any defini-
tion of class which does not include a comparison with other
economic groups, i.e., absent relational comparisons. How-
ever, gender and class identities are not the only types that
have relational content: many ethnic and religious identities
also include relational comparisons, however the comparisons
are usually targeted, e.g., Chinese may compare themselves to
Japanese or Koreans in terms of defining who they are, but not
to Mexicans or Algerians.
Qualitative & Multi-Method Research, Spring 2009
Relational content is also crucial for social identity theory
(SIT), which hypothesizes that the creation of an in-group
identity will tend to produce competitive behavior with out-
groups, because the process of in-group identity creation by
necessity requires, or leads to, the devaluation of out-groups
(Tajfel 1981). In social identity theory, the central causal pro-
cess in behavior derives from in-group and out-group differ-
entiation, not the roles or identity traits per se that are attrib-
uted to in-groups and out-groups. In this case, action is in
some sense a reaction to, and conditioned by the existence of,
those who are different. Some relationships (those with groups
socially recognized as similar) will be more cooperative than
others (those with groups recognized as different) even if the
same issue is at stake (such as territory, power, or status). SIT
is one of the most well-established research programs in social
psychology and it provides a clear mechanism linking identi-
ties and behavior.
Cognitive Models
Finally, the content of an identity can come in the form of
cognitive models. In the broadest sense, a cognitive model
may be thought of as a worldview, or a framework that allows
members of a group to make sense of social, political, and
economic conditions.3 The cognitive content of a collective
identity describes how group membership is associated with
explanations of how the world works as well as descriptions of
the social reality of the group—a group’s ontology and episte-
mology. For example, according to Brubaker, Loveman, and
Stamatov, “what cognitive perspectives suggest, in short, is
that race, ethnicity, and nation are not things in the world but
ways of seeing the world. They are ways of understanding
and identifying oneself, making sense of one’s problems and
predicaments, identifying one’s interests, and orienting one’s
action” (2004: 47).
Cognitive content, rather than implying an identity-based
theory of action (à la norms or SIT), implies a theory of iden-
tity-based interpretation and development of interests (which
of course may be closely related to action). Thus, attention to
cognitive models suggests how identity affects the way ac-
tors understand the world and how their preferences regard-
ing material or social action can be influenced by their identi-
ties.
Contestation
The content or collective meaning of identities is neither
fixed nor predetermined. Rather, content is the outcome of a
process of social contestation within the group. Indeed, much
of identity discourse is the working out of the meaning of a
particular collective identity through the contestation of its
members. Individuals are continuously proposing and shap-
ing the meanings of the groups to which they belong, hence
content is always contested and contestation is the key to
understanding how the meaning or content of identities change
over time. For example, Episcopalian congregants debate
whether to accept gay priests, Israelis debate the acceptability
of territorial boundaries, etc.
Specific interpretations of the meaning of an identity are
43
Qualitative & Multi-Method Research, Spring 2009
sometimes widely shared among members of a group and some-
times less widely shared. At a minimum, contestation can be
thought of as a matter of degree—the content of collective
identities can be more or less contested. Indeed, the further
apart the contending interpretations of a collective identity
prove to be, the more that identity will be fragmented into
conflicting and potentially inconsistent understandings of what
the group’s purposes or relations should be.
Because the content of an identity is the product of con-
testation, the very data that a scholar extracts from a group
elucidate, in manner and degree, the members’ consensus and
disagreement about constitutive norms, consensus and con-
gruence of the social purposes ascribed to an identity, agree-
ment about meanings attached to out-groups, and coherence
of shared cognitive models. By treating contestation as an
empirical question, one can take snapshots of the degree of
stability or flux in identities as they evolve, as they are chal-
lenged, and as they are constructed and reconstructed, hence
addressing the theoretical tension between measurement and
the fluidity of social identities. In addition, a focus on contes-
tation allows one to examine the strategic process of identity
construction including issues of authenticity, internalization,
etc.
Finally, by considering the level of contestation regarding
each type of content within identities, one arrives at a neces-
sarily constructivist approach to identity without having to
assume that actors on the ground view their identities as con-
structed. Where there is little contestation, one might con-
clude that that part of identity content is taken for granted or
considered “natural.” Thus, one can appreciate some appar-
ently “primordial” aspects of identity without taking a pri-
mordialist theoretical stance that denies the possibility for con-
testation at different times and places.
Methods for Measuring Identity Content and Contestation
In our survey of the scholarly literature on identity, we
found that surveys, content analysis, discourse analysis, and
ethnography were the most widely used methods. We did not
discover any systematic links between these methods and the
types of identity or the types of content they were used to
measure, although nearly all studies of identity included some
sort of case study. We also identified two additional meth-
ods—cognitive mapping and experiments—that offer great
promise to supplement the dominant methods. In order to high-
light the variety of methods available to identity researchers,
below we briefly describe some of the ongoing work on mea-
surement of identities that appears in our recently published
edited volume, Measuring Identity: A Guide for Social Scien-
tists (Cambridge University Press, 2009).
Part I of the volume considers identity definition, con-
ceptualization, and measurement alternatives. The first chap-
ter is an extended discussion of our analytical framework for
measuring identity content and contestation described above.
The next two chapters demonstrate how various methods can
be combined to define and measure identity. Chapter 2, by
Henry E. Brady and Cynthia S. Kaplan, considers the concept-
ualization and measurement of politically relevant social iden-
44
tities in four Soviet republics, using several different methods
and types of data—history, demography, surveys, primary
source materials, content analysis, and discourse analysis—
in order to measure the political role of ethnicity at the mass
public and elite level. Similarly, Chapter 3, by Donald A. Sylvan
and Amanda K. Metskas, considers a range of methodological
options, but in this chapter the focus is on Israeli-Palestinian
relations. Based on five research projects that used experimen-
tal, survey, interview, narrative, and text-based data, Sylvan
and Metskas consider the trade-offs between alternative ap-
proaches to measuring identity in this context.
The remaining chapters all proceed by presenting a mea-
surement method that the authors have applied to a particular
empirical issue. In essence, these chapters represent extended
methodological discussions based on individual authors’ ex-
periences using identity as a variable in larger research projects.
Each of these chapters details the workings of a particular
definition and method, discussing both its advantages and
disadvantages. The chapters are divided into four sections:
surveys, content analysis (including cognitive mapping), dis-
course analysis and ethnography, and experiments.
In the section on surveys, Chapter 4, by Taeku Lee, exam-
ines the conspicuous gap between social theory on race and
ethnicity, which stresses its fluidity, multiplicity, and contin-
gency, and quantitative, survey-based studies of race and
ethnicity, which remain focused on finding a common, fixed set
of categories that reliably and validly reflect how individuals
think of themselves in racial or ethnic terms. This chapter pro-
poses a new approach to measuring ethnoracial self-identifi-
cation. Lee’s “identity point allocation” method gives respon-
dents latitude over how many groups to identify with and how
to weight the strength of their identification with each group.
Chapter 5, by Jack Citrin and David O. Sears, explores the
implications of holding multiple identities, concentrating on
how individuals balance national and ethnic identities in
multiethnic states. Citrin and Sears’s case study considers the
United States and the current demographic and ideological
challenge to the idea of E pluribus unum. By reviewing alter-
native measures of identity and some of the obstacles to sys-
tematic measurement, the chapter argues for the need to build
upon more qualitative explorations of the content of identities
before undertaking survey research. It then uses survey re-
search—both national samples for representativeness and
pooled data from Los Angeles—to explore how citizens con-
ceive of and prioritize their national and ethnic identities. In
Chapter 6, Michael Dawson explores the concept of racial iden-
tity. Much of the work on black political identity, and increas-
ingly work focused on Latino/a and Asian American political
identity as well, has focused on one particular construction of
racial identity—that of “linked fate.” Empirically, Dawson tests
the degree to which racial identity, operationalized by the con-
cept of “linked fate,” is still able to shape African Americans’
political beliefs.
The next section contains three chapters that use content
analysis to measure identity. In Chapter 7, Kimberly A.
Neuendorf and Paul D. Skalski consider content analysis in
terms of a quantitative investigation based on the coding of

message characteristics. They present three main coding pos-
sibilities: human coding from a preset scheme; computer text
coding from a preset scheme; and “emergent” computer text
coding, where dimensions derive from the data at hand. The
chapter examines each approach using original identity-based
data as well as data taken from the research of other identity
scholars, and the advantages and limitations of the various
approaches are considered. Chapter 8, by Robalyn Stone and
Michael Young, examines computer-generated cognitive maps
as a tool for extracting information about identity concepts in
texts. Cognitive maps are a representation of the fundamental
underlying belief system expressed in a text. The chapter uses
two unique software programs: Profiler Plus and Worldview.
Stone and Young show how these applications can be used to
examine how different Iraqi leaders conceive of the main traits
and characteristics of different ethnic and religious identities,
illuminating the conceptual distance across individuals within
identity groups. Chapter 9, by Kanchan Chandra, has been
outlined to some extent in her contribution to this symposium.
In the chapter, she discusses an important new constructivist
dataset on several concepts related to ethnic identity and in-
stitutions—CDEI (Constructivist Dataset on Ethnicity and In-
stitutions) including EVOTE, which is the focus of the chapter.
The next two chapters illustrate how discourse analysis
and ethnography can illuminate the measurement of identity
content. Chapter 10 discusses a constructivist theory of iden-
tity that is at once social, structural, and cognitive. The author,
Ted Hopf, invokes discourse analysis to explore three logics
of social order—consequentialism, appropriateness, and
habit—and to relate them to the concept of identity. He then
applies this theory to the study of a state’s foreign policy
choices, and in particular, the case of the Sino-Soviet split.
Chapter 11, by Laura Adams, reviews the way that identity has
been measured in a select but diverse group of ethnographic
studies, focused mainly on post-Soviet Uzbekistan. The chap-
ter outlines the strengths and weaknesses of ethnographic
methods and explores the ways that ethnographers deal with
the challenges of their research process.
Finally, the last chapter explores the possibility of em-
ploying the experimental method to measure social identities.
In Chapter 13, Rose McDermott provides an overview of the
experimental literature on identity research. This chapter be-
gins with a substantive discussion of the experimental work
that has been conducted on social identity, which as it turns
out is concerned almost exclusively with the ramifications, im-
plications, and limitations of social identity theory. The chap-
ter discusses limitations of existing experimental work for ap-
plications to political science, and examines some of the ways
in which the method of experimentation might be expanded to
investigate other realms of social identity in political contexts.
Conclusions
Identity content, and the social contestation that produces
and follows it, we argue is a critical part of measuring ethnic
and other social identities. It is by disaggregating and examin-
ing the content of identity that we can make comparisons across
identity types. Moreover, identity content helps us understand
Qualitative & Multi-Method Research, Spring 2009
the mechanisms by which having an identity shapes interests
and actions; in particular, we argue that constitutive norms
and relational comparisons suggest mechanisms for identity-
based behavior, and social purposes and cognitive content
address the formation of identity-based interests.
Finally, we want to emphasize the variety of methodologi-
cal choices available to scholars interested in measuring iden-
tity content and making comparisons across identity types. In
our edited volume we sought to include several different meth-
odologies on a wide range of empirical topics across political
science. We hope that this collection moves the debate on
measuring the content of social identities forward by suggest-
ing that methodological rigor need not imply the use of a single
method or definition of identity. By examining identity content
using diverse research strategies, we hope to get a step closer
to explaining the power of identity as a variable.
Notes
1
This article largely summarizes our recent work; see Abdelal et al.
(2006, 2009).
2
In Chandra’s terminology, these different types of identities rep-
resent different “dimensions.”
3
Some use other terms besides “worldviews.” Denzau and North
(1994), for example, use “shared mental models.”
References
Abdelal, Rawi, Yoshiko Herrera, Iain Johnston, and Rose McDermott.
2006. “Identity as a Variable.” Perspectives on Politics 4:4, 695–
711.
Abdelal, Rawi, Yoshiko Herrera, Alastair Iain Johnston, and Rose
McDermott, eds. 2009. Measuring Identity: A Guide for Social
Scientists. New York: Cambridge University Press.
Alesina, Alberto, Arnaud Devleeschauwer, William Easterly, Sergio
Kurlat, and Romain Wacziarg. 2003. “Fractionalization.” Journal
of Economic Growth 8:2, 155–194.
Brubaker, Rogers and Frederick Cooper. 2000. “Beyond ‘Identity.’”
Theory and Society 29:1, 1–47.
Brubaker, Rogers, Maria Loveman, and Peter Stamatov. 2004.
“Ethnicity as Cognition.” Theory and Society 33:1, 31–64.
Chandra, Kanchan. 2006. “What is Ethnic Identity and Does it Mat-
ter?” Annual Review of Political Science, Vol. 9, 397–424.
Denzau, Arthur T. and Douglass C. North. 1994. “Shared Mental
Models: Ideologies and Institutions.” Kyklos 47:1, 3–31.
Horowitz, Donald. 1985. Ethnic Groups in Conflict. Berkeley: Uni-
versity of California Press.
Jepperson, Ronald L., Alexander Wendt, and Peter J. Katzenstein.
1996. “Norms, Identity, and Culture in National Security.” In The
Culture of National Security: Norms and Identity in World Politics,
Peter J. Katzenstein, ed. New York: Columbia University Press,
33–75.
Tajfel, Henri. 1981. Human Groups and Social Categories: Studies in
Social Psychology. London: Cambridge University Press.
45
Qualitative & Multi-Method Research, Spring 2009
Book Notes
Book descriptions are excerpted from publisher’s websites.
If you would like to recommend a book to be included in
this section, email Joshua C. Yesnowitz, the assistant
editor of QMMR, at jcyesnow@bu.edu.
Abdelai, Rawi, Yoshiko M. Herrera, Alastair Iain Johnston, and
Rose McDermott, eds. 2009. Measuring Identity: A Guide
for Social Scientists. New York: Cambridge University Press.
The concept of identity has become increasingly prominent in the
social sciences and humanities. Analysis of the development of social
identities is an important focus of scholarly research, and scholars
using social identities as the building blocks of social, political, and
economic life have attempted to account for a number of discrete
outcomes by treating identities as causal factors. The dominant im-
plication of the vast literature on identity is that social identities are
among the most important social facts of the world in which we live.
Abdelal, Herrera, Johnston, and McDermott have brought together
leading scholars from a variety of disciplines to consider the concep-
tual and methodological challenges associated with treating identity
as a variable, offer a synthetic theoretical framework, and demon-
strate the possibilities offered by various methods of measurement.
The book represents a collection of empirically-grounded theoretical
discussions of a range of methodological techniques for the study of
identities.
Byrne, David and Charles Ragin, eds. 2009. The Sage Hand-
book of Case-Based Methods. Newbury Park, CA: Sage Pub-
lications.
This book provides a fresh and stimulating approach to causal analy-
sis in the social sciences. International experts provide not just the
philosophical arguments for a case-based approach to research but
also detailed chapters on “why to,” “when to” and “how to.” Tradi-
tional distinctions between qualitative and quantitative are rejected in
favor of a case-based approach which is applicable across the social
sciences and beyond.
Caramani, Daniele. 2008. Introduction to the Comparative
Method with Boolean Algebra. Newbury Park, CA: Sage
Publications.
Offering students and researchers in the behavioral and social sci-
ences a brief and accessible introduction to the comparative method,
it is ideal for students of public administration, policy, sociology,
political science, social psychology, and international relations. It
provides readers with basic guidelines for comparative research by
addressing all key methodological issues.
Collier, David and John Gerring, eds. 2009. Concepts and Meth-
ods in the Social Sciences: The Tradition of Giovanni Sar-
tori. London: Routledge.
Careful work with concepts is a cornerstone of good social science
methodology. Concepts and Method in Social Science demonstrates
the crucial role of concepts, providing a timely contribution that
draws both on the classic work of Giovanni Sartori and the writing of
a younger generation of scholars. In this volume, major writings of
Sartori are juxtaposed with other work that exemplifies important
46
approaches to concept analysis. The book is organized into three key
sections: Part I: Sartori on Concepts and Methods—including an
examination of the necessary logical steps in moving from concep-
tual-ization to measurement and the relationships among meanings,
terms and observations; Part II: Extending the Sartori Tradition—
eminent scholars analyze five key ideas in concept analysis: revolu-
tion, culture, democracy, peasants and institutionalization within the
context of the Sartori tradition; Part III: In the Academy and Be-
yond—both an engaging autobiographical essay written by Giovanni
Sartori and reflections from former students provide a unique context
in which to situate this varied and rigorous discussion of concept
analysis and qualitative methods. Concepts and Method in Social
Science is an accessible text that is well suited to advanced under-
graduates and graduate students, providing a distinct and coherent
introduction to comparative political analysis.
Krook, Mona Lena. 2009. Quotas for Women in Politics: Gen-
der and Candidate Selection Reform Worldwide. Oxford:
Oxford University Press.
In recent years, political parties and national legislatures in more than
100 countries have adopted quotas for the selection of female candi-
dates to political office. Despite the rapid diffusion of these measures
around the globe, most research has focused on single countries—or,
at most, the presence of quotas within one world region. Due to
limited evidence, explanations for adoption and impact of gender
quotas derived from one study frequently contradict findings from
other cases. Quotas for Women in Politics is the first book to address
quotas as a global phenomenon in order to provide greater analytical
leverage in explaining their spread and impact in diverse contexts
around the world. It is organized around two sets of questions: First,
why are quotas adopted? Which actors are involved in quota cam-
paigns, and why do they support or oppose quota measures? Sec-
ond, what effects do quotas have on existing patterns of political
representation? Are these provisions sufficient for bringing more
women into politics? Or does their impact depend on other features
of the broader political context? Synthesizing the literature on quota
policies, Mona Lena Krook develops a framework for analyzing the
spread of quota provisions and the reasons for variations in their
effects. She then uses this framework to examine and compare differ-
ent types of quota policies in Pakistan and India; Sweden and the
United Kingdom; and Argentina and France.
Landman, Todd and Edzia Carvalho. 2009. Measuring Human
Rights. London: Routledge.
The measurement of human rights has long been debated within the
various academic disciplines that focus on human rights, as well as
within the larger international community of practitioners working in
the field of human rights. Written by leading experts in the field, this
is the most up-to-date and comprehensive book on how to measure
human rights. Measuring Human Rights draws explicitly on the in-
ternational law of human rights to derive the content of human rights
that ought to be measured; contains a comprehensive methodological
framework for operationalizing this human rights content into human
rights measures; includes separate chapters on the methods, strengths,
and biases of different human rights measures, including events-based,
standards-based, survey-based, and socio-economic and administra-
tive statistics; covers measures of civil, political, economic, social and
cultural rights; and includes a complete bibliography, as well as sources
and locations for datasets useful for the measurement of human rights.
This volume offers a significant and timely addition to this important
area of work in the field of human rights, and will be of interest to
academics and NGOs, INGOs, international governmental organiza-

tions, international financial institutions, and national governments
themselves.
Munck, Gerardo L. 2009. Measuring Democracy: A Bridge
between Scholarship and Politics. Baltimore: Johns Hopkins
University Press.
Although democracy is a widely held value, concrete measurement of
it is elusive. Gerardo L. Munck’s constructive assessment of the
methods used to measure democracies promises to bring order to the
debate in academia and in practice. Drawing on his years of academic
research on democracy and measurement and his practical experience
evaluating democratic practices for the United Nations and the Orga-
nization of American States, Munck’s discussion bridges the theories
of academia with practical applications. In proposing a more open
and collaborative relationship between theory and action, he makes
the case for reassessing how democracy is measured and encourages
fundamental changes in methodology. Munck’s field-tested frame-
work for quantifying and qualifying democracy is built around two
instruments he developed: the UN Development Programme’s Elec-
toral Democracy Index and a case-by-case election-monitoring tool
used by the OAS. Measuring Democracy offers specific, real-world
lessons that scholars and practitioners can use to improve the quality
and utility of data about democracy.
Taagepera, Rein. 2008. Making Social Science More Scien-
tific: The Need for Predictive Models. Oxford: Oxford Uni-
versity Press.
In his challenging new book Rein Taagepera argues that society needs
more from social sciences than they have delivered. One reason for
falling short is that social sciences have depended excessively on
regression and other statistical approaches, neglecting logical model
building. Science is not only about the empirical “What is?” but also
very much about the conceptual “How should it be on logical grounds?”
Statistical approaches are essentially descriptive, while quantitatively
formulated logical models are predictive in an explanatory way. Mak-
ing Social Sciences More Scientific contrasts the predominance of
statistics in today’s social sciences and predominance of quantita-
tively predictive logical models in physics. It shows how to con-
struct predictive models and gives social science examples. Making
Social Sciences More Scientific is useful to students who wish to learn
the basics of the scientific method and to all those researchers who
look for ways to do better social science.
Article Notes
Bevir, Mark and Asaf Kedar. 2008. “Concept Formation in
Political Science: An Anti-Naturalist Critique of Qualitative
Methodology.” Perspectives on Politics 6:3, 503–517.
This article offers an anti-naturalist philosophical critique of the natu-
ralist tendencies within qualitative concept formation as developed
most prominently by Giovanni Sartori and David Collier. We begin
by articulating the philosophical distinction between naturalism and
anti-naturalism. Whereas naturalism assumes that the study of hu-
man life is not essentially different from the study of natural phe-
nomena, anti-naturalism highlights the meaningful and contingent na-
ture of social life, the situatedness of the scholar, and so the dialogical
nature of social science. These two contrasting philosophical ap-
proaches inspire, in turn, different strategies of concept formation.
Naturalism encourages concept formation that involves reification,
essentialism, and an instrumentalist view of language. Anti-natural-
Qualitative & Multi-Method Research, Spring 2009
ism, conversely, challenges reified concepts for eliding the place of
meanings, essentialist concepts for eliding the place of contingency,
and linguistic instrumentalism for eliding the situatedness of the scholar
and the dialogical nature of social science. Based on this philosophical
framework, we subject qualitative concept formation to a philosophical
critique. We show how the conceptual strategies developed by Sartori
and Collier embody a reification, essentialism, and instrumentalist
view of language associated with naturalism. Although Collier’s work
on concept formation is much more flexible and nuanced than Sartori’s,
it too remains attached to a discredited naturalism.
Dunning, Thad. 2008. “Improving Causal Inference: Strengths
and Limitations of Natural Experiments.” Political Research
Quarterly 61:2, 282–293.
Social scientists increasingly exploit natural experiments in their re-
search. This article surveys recent applications in political science,
with the goal of illustrating the inferential advantages provided by
this research design. When treatment assignment is less than “as if”
random, studies may be something less than natural experiments, and
familiar threats to valid causal inference in observational settings can
arise. The author proposes a continuum of plausibility for natural
experiments, defined by the extent to which treatment assignment is
plausibly “as if” random, and locates several leading studies along this
continuum.
Lambach, Daniel and Dragan Gamberger. 2008. “Temporal
Analysis of Political Instability Through Descriptive Sub-
group Discovery.” Conflict Management and Peace Sci-
ence 25:1, 19–32.
This paper analyzes the Political Instability Task Force (PITF) dataset
using a new methodology based on machine learning tools for sub-
group discovery. While the PITF used static data, this study em-
ploys both static and dynamic descriptors covering the five-year
period before onset. The methodology provides several descriptive
models of countries especially prone to political instability. For the
most part, these models corroborate the PITF’s findings and support
earlier theoretical works. The paper also shows the value of subgroup
discovery as a tool for developing a unified concept of political insta-
bility as well as for similar research designs.
Mahoney, James, Erin Kimball, and Kendra L. Koivu. 2009.
“The Logic of Historical Explanation in the Social Sciences.”
Comparative Political Studies 42:1, 114–146.
Historical explanations seek to identify the causes of outcomes in
particular cases. Although social scientists commonly develop his-
torical explanations, they lack criteria for distinguishing different types
of causes and for evaluating the relative importance of alternative
causes of the same outcome. This article first provides an inventory
of the five types of causes that are normally used in historical expla-
nations: (1) necessary but not sufficient, (2) sufficient but not neces-
sary, (3) necessary and sufficient, (4) INUS, and (5) SUIN causes. It
then introduces a new method—sequence elaboration—for evaluat-
ing the relative importance of causes. Sequence elaboration assesses
the importance of causes through consideration of their position within
a sequence and through consideration of the types of causes that
make up the sequence as a whole. Throughout the article, method-
ological points are illustrated with substantive examples from the
field of international and comparative studies.
47
Qualitative & Multi-Method Research, Spring 2009
McAdam, Doug, Sidney Tarrow, and Charles Tilly. 2008. “Meth-
ods for Measuring Mechanisms of Contention.” Qualita-
tive Sociology 31:4, 307–331.
A substantial intellectual movement has been growing in the social
sciences around the adoption of mechanism- and process-based ex-
planations as complements to variable-based explanations, or even as
substitutes for them. But once we have recognized the validity and
dignity of studying mechanisms and processes, what is the next step?
Recently, both political scientists’ and sociologists’ discussions have
begun to turn away from correlation to mechanism-based approaches
to causation. But there is still a widespread assumption that mecha-
nisms are unobservable. We maintain that ways can be developed to
observe the presence or absence of mechanisms either directly or
indirectly. In this paper, by way of example, we put forward four
methods—two direct and two indirect—for measuring mechanisms
of contention.
Narang, Vipin and Rebecca M. Nelson. 2009. “Who Are These
Belligerent Democratizers? Reassessing the Impact of De-
mocratization on War.” International Organization 63, 357–
379.
In a key finding in the democratic peace literature, Mansfield and
Snyder argue that states with weak institutions undergoing incom-
plete transitions to democracy are more likely to initiate an external
war than other types of states. We show that the empirical data do
not support this claim. We find a dearth of observations wherein
complete democratizers with weak institutions participated in war.
Additionally, we find that the statistical relationship between incom-
plete democratization and war is entirely dependent on the dismem-
berment of the Ottoman Empire prior to World War I. We also find
that the case selection in Mansfield and Snyder rarely involved in-
complete democratizers with weak institutions. We therefore con-
clude that the finding that incomplete democratizers with weak insti-
tutions are more likely to initiate or participate in war is not sup-
ported by the empirical data.
Rohlfing, Ingo. 2008. “What You See and What You Get: Pit-
falls and Principles of Nested Analysis in Comparative Re-
search.” Comparative Political Studies 41:11, 1492–1514.
In a recent contribution to this journal, Munck and Snyder found that
many studies suffer from a deficient application of qualitative and
quantitative methods. They argue that the combination of small-n and
large-n analysis represents a viable method for promoting the produc-
tion of knowledge. Recently, Evan Lieberman proposed nested analy-
sis as a rigorous approach for comparative research that builds on the
complementary strengths of quantitative and qualitative analysis. In
this paper, the author examines the methodological potential of nested
inference to advance comparative political analysis, arguing that the
specific methodological problems of nested designs have not been
fully appreciated. It is shown that, under certain circumstances, noth-
ing is gained from a nested analysis. On the contrary, one might lose
more than one gains compared to single-method designs. The author
suggests specific methodological principles that take these problems
into account to make nested analysis fruitful for comparative studies.
48
Announcements
APSA Panels/Roundtables Created (or Co-Organized) by
Division 46: Qualitative and Multi-Method Research
September 3–September 6, 2009, Toronto, ON, Canada
Complexity and Interdependence in World Politics:
New Qualitative Approaches
Chair: Gary Goertz, University of Arizona
Participants:
Gary Goertz, University of Arizona: “The Political and Institutional
Construction of International Regions: Conceptualization and
Operationaliation.”
Daniel J. Levine, Johns Hopkins University: “Holes in the Whole:
Negative Dialectics and the Limits of Integration Theory.”
Vsevolod Gunitskiy, Columbia University: “A New Path or a Cul-
de-sac? Complex Adaptive Systems and International Relations
Theory.”
Tanja Pritzlaff, University of Bremen, Germany: “Complexity and
Stability in Contexts of Joint Decision-Making: An Experimental
Study.”
Charles L. Mitchell, Grambling State University: “Implications of
Qualititative Methods for Studying International Politics.”
Discussant: Kaija Schilde, University of Pennsylvania
Understanding Experiences Across the Subfields:
Rhetoric, Phenomenology, Fieldwork, Framing/Narratives,
and Textual Ethnography
Chair: Dvora Yanow, Vrije Universiteit, Amsterdam
Participants:
Dennis C. Galvan and Gerald Berk, University of Oregon: “How to
Research Institutions Experientially: Methods of Creative
Syncretism.”
Xymena Kurowska, Central European University: “Embedded IR-
ist: Learning about the EU’s Practices in External Assistance.”
Johannes Morrow, SUNY-Albany: “Methodological Issues in
Comparative Political Theory: Perspectives from Indigenous
Studies.”
Nick Turnbull, University of Manchester, United Kingdom: “The
Rhetorical Analysis of Politics.”
Cyrus Ernesto Zirakzadeh, University of Connecticut: “Frames
and Narratives: Two Modes of Political Understanding; Two
Forms of Scholarly Interpretation.”
Discussants: Peregrine Schwartz-Shea, University of Utah;
Timothy V. Kaufman-Osborn, Whitman College
Debating Research Designs: Do Qualitative and Interpretive
Logics of Inquiry Differ? Should They?
Chair: Dvora Yanow, Vrije Universiteit, Amsterdam
Participants: Colin Elman, Syracuse University; John Gerring,
Boston University; Julie L. Novkov, SUNY-Albany; Sanford F.
Schram, Bryn Mawr College; Peregrine Schwartz-Shea, Uni-
versity of Utah
The Methods Café
Chair: Dvora Yanow, Vrije Universiteit, Amsterdam
Participants:
Lisa Wedeen, University of Chicago: “Critical Constructivist and
Discourse Analysis.”
Katherine Cramer Walsh, University of Wisconsin, Madison: “Field
 Qualitative & Multi-Method Research, Spring 2009
Research I (Participant Observation; Political Ethnography): United The Epistemological Foundations of Mixed-Method Research
States.”
Chair: Amel F. Ahmed, University of Massachusetts-Amherst
Jan Kubik, Rutgers University: “Field Research II (Political Ethnog-
Participants:
raphy, Participant Observation): Overseas.”
David Kuehn and Ingo Rohlfing, University of Heidelberg,
Peregrine Schwartz-Shea, University of Utah: “Generalizing? Valid-
Germany: “Are We Really Bridging the Gap? On the Conflicting
ity? Reliability?”
Epistemological Foundations of Multi-Method Research.”
Frederic C. Schaffer, University of Massachusetts-Amherst: “Con-
Abhishek Chatterjee, University of Virginia: “Ontology, Epistemol-
versational and Ordinary Language Interviewing.”
ogy, and Multiple Methods.”
Amel F. Ahmed, University of Massachusetts-Amherst: “Mapping
Illiberal Politics in Liberal States:
the Epistemological Commitments of Methods: A Framework
Studying the “Rough Edges of Democracy”
for Mixed-Method Research.”
Chair: Martha Crenshaw, Stanford University Discussants: Jeffrey T. Checkel, Simon Fraser University, Canada;
Participants: Ted Hopf, Ohio State University
Giovanni Capoccia, Oxford University, United Kingdom: “Demo-
cratic Signalling and Restrictions to Pluralism: Banning Extremist Process Tracing in International and Comparative Politics:
Parties in Advanced Democracies.” Achievements and Challenges
Terri E. Givens, University of Texas-Austin: “Antidiscrimination
Chair: Jeffrey T. Checkel, Simon Fraser University, Canada
Policy and Rights: Majority vs. Minority.”
Participants: Andrew Bennett, Georgetown University; James A.
Jonathan A. Laurence, Boston College: “Making Islam Safe for
Caporaso, University of Washington; Colin Elman, Syracuse
Democracy: Legal Restrictions on Political Islamist Federations
University; James Mahoney, Northwestern University; Vincent
in Western European Democracies.”
Pouliot, McGill University, Canada
Ami Pedahzur and Eran Zaidise, University of Texas-Austin:
“Weak Democratic States and Reactions to Extremism.”
Virtues and Limits of Mixed-Method
Christian Davenport, University of Maryland: “When Democra-
Research in Diverse Contexts
cies Kill. A Comparative Study of the USA, Northern Ireland,
Rwanda, and India.” Chair: Aaron Schneider, Tulane University
Discussant: Martha Crenshaw, Stanford University Participants:
Gitte Sommer Harrits, University of Aarhus, Denmark: “Under-
Focus on Metaphor: New Perspectives standing Social and Political Practice: A Mixed-Method
on Language and Discourse Strategy.”
Aaron Schneider, Tulane University: “Change and Complexity in
Chair: Terrell Carver, University of Bristol, United Kingdom
Stateness: Quantitative, Qualitative, and Interpretive Tools to
Participants:
Make Sense of State Authority in an Age of Globalization.”
Takashi Shogimen, University of Otago, New Zealand: “Context
Ariel Ahram, University of Oklahoma: “Conceptual Stretching in
and Metaphor: A New Approach to the History of Ideas.”
Mixed-Method Research.”
Jernej Pikalo, University of Ljubljana, Slovenia: “Analyzing the
Kaija Schilde, University of Pennsylvania: “Triangulating Methods
Interrelationship between Metaphors and Contexts: Informing
to Assess the Performance of International Bureaucracies: An
and Performing.”
Analysis of EU Institutions Through Case Studies and Surveys.”
Dag Stenvoll, University of Bergen, Norway: “Using Metaphors to
Discussant: Herlin Chien, National Sun Yat-sen University, Taiwan
Analyze the US Report on ‘Trafficking in Persons.’”
Steffen G. Schneider and Frank Nullmeier, University of Bremen,
Qualitative Research in Post-Communist Space
Germany: “Metaphorical Concepts and the Discursive Con-
struction of Legitimacy: The Framing of (Inter-)National Gover- Chair: Jessica Allina-Pisano, University of Ottawa, Canada
nance Arrangements in Media Discourses.” Participants:
Discussant: Veronique Mottier, University of Lausanne, Switzer- Paul Goode, University of Oklahoma: “Redefining Russia: Quali-
land tative Research and Western Political Science.”
Jessica Allina-Pisano and Andre Simonyi, University of Ottawa,
Ethnographic Methods in Political Science: Canada: “Power, Space, and Movement in the Eastern Border-
What Difference Can They Make? lands of the European Union.”
Andre Simonyi, University of Ottawa, Canada: “Beyond ‘Beyond
Chair: Edward Schatz, University of Toronto-Mississauga, Canada
Identity’: The Creation of Magyar and Korean ‘Minorities’ in
Participants: Calvin Chen, Mount Holyoke College; Jan Kubik;
Ukraine.”
Rutgers University-New Brunswick; Timothy Pachirat, New
Daniel J. Beers, Indiana University: “Sensitive Questions Demand
School University; Dorian T. Warren, Columbia University;
Sensitive Methods: A Comparative Analysis of Interview and
Edward Schatz, University of Toronto-Mississauga, Canada
Survey Responses to Questions about Corruption and Profes-
Discussant: Dvora Yanow, Vrije Universiteit, Amsterdam
sional Misconduct in the Romanian Judiciary.”
Discussant: Paul Goode, University of Oklahoma
Is There a Multimethod Consensus in Comparative Politics?
Chair: Rudra Sil, University of Pennsylvania Statistical Models and Causal Inference:
Participants: Michael J. Coppedge, University of Notre Dame; David Freedman’s Dialogue with the Social Sciences
Yoshiko M. Herrera, University of Wisconsin-Madison; Evan S.
Chair: Jasjeet Singh Sekhon, University of California-Berkeley
Lieberman, Princeton University; Amel F. Ahmed, University of
Participants: Jason Seawright, Northwestern University; Donald P.
Massachusetts-Amherst; Dan Slater, University of Chicago
Green, Yale University; Henry E. Brady, University of Califor-

49
Qualitative & Multi-Method Research, Spring 2009
nia-Berkeley; Wendy K. Tam Cho, University of Illinois-
Urbana-Champaign; Thad Dunning, Yale University
Discussant(s): Jasjeet Singh Sekhon, University of California-
Berkeley; David Collier, University of California-Berkeley
History, Identity, Political Violence:
The Relative Merits of Qualitative Methods to Explain
Complex and Dynamic Phenomena
Chair: Jonathan Githens-Mazer, University of Exeter, United
Kingdom
Participants:
Jonathan Githens-Mazer, University of Exeter, United Kingdom:
“Causal Processes, Radicalisation and Bad Policy: The Impor-
tance of Case Studies of Radical Violent Takfiri Jihadism for
Establishing Logical Causality.”
Orla Lynch, University of St. Andrews, Scotland: “The Advantages
of Qualitative Methods in Difficult-to-Research Subject Pop-
ulations: Triangulating Interviews and Secondary Sources.”
Matthew Goodwin, University of Manchester, United Kingdom:
“Beyond a Snapshot Approach: Findings from Life-History
Interviews with Extreme Right-Wing Activists.”
Robert A. Lambert, University of St. Andrews, Scotland: “Under-
standing Muslim Community Perspectives of Violent Extrem-
ism: A Qualitative Case Study in London.”
Discussant: Basia Spalek, University of Birmingham, United
Kingdom
Taking Research Design Seriously in
Ideational Approaches to International Relations
Chair: Jeffrey W. Legro, University of Virginia
Participants:
Joshua Busby and Jonathan Monten, University of Texas-Austin:
“When John Bolton is the Life of the Party: Explaining the
Erosion of Multilateralism in the GOP.”
Jennifer L. Erickson, Cornell University: “Reputation and Image in
International Institutions: International Pressures to Adopt
‘Responsible’ Arms Transfer Policy.”
Andrew Yeo, Catholic University of America: “Contestation or
Consensus? Ideas, Foreign Policy Beliefs, and U.S. Alliance
Relations.”
Stephanie Claudia Hofmann, Cornell University: “European
Security in the Shadow of NATO: Party Ideology and Institu-
tion Building.”
Stephen Craig Nelson and Andrew Yeo, Cornell University:
“Methodological Challenges and Progress in Ideational Re-
search.”
Discussant: Kathleen R. McNamara, Georgetown University
Case Study Meta-Analysis: Methodological Challenges and
Applications in Political Science
Chair: Jens Newig, Leuphana Universität Lüneburg, Germany
Participants:
Jens Newig and Oliver Fritsch, Leuphana Universität Lüneburg,
Germany: “Does Participatory Governance Lead to Better
Environmental Outcomes? Methodology and Results from a
Transatlantic Comparative Meta-Analysis of 60 Case Studies in
Environmental Decision Making.”
Claudio M. Radaelli and Theofanis Exadaktylos, University of
Exeter, United Kingdom: “Research Design and Causal Analysis
in European Studies. A Meta-Analysis of the Europeanization
Literature.”
Jason Jensen, University of North Dakota: “Cumulating the
50
Results from Political Science, Public Administration and Public
Policy Case Studies Using Meta-Analysis: Issues, Examples,
and Recommendations.”
Kurt Gaubatz and Katya Drozdova, Stanford University and
National Security Innovations (NSI): “Focusing on the Structure
of Uncertainty: Using Information Models to Enhance the
Structured-Focused Case Study Method.”
Discussant: Claudio M. Radaelli, University of Exeter, United
Kingdom
Qualitative Approaches to Institutional and
Policy Change in American Politics
Chair: Donald Rosdil, Northern Virginia Community College
Participants:
Morris D. Bidjerano, SUNY-Albany: “Complexifying Collabora-
tion.”
Adam Myers and Curtis Nichols, University of Texas-Austin:
“New Insights About Critical Junctures: Lessons From The
Study of Governing-Majority Formation in American Politics.”
Alex Leland Medler, University of Colorado-Boulder: “Mapping
Complex Coalitions: Using Frames and Policy Positions to
Identify Ideologues, Pragmatists, and Dogmatic Coalition Mem-
bers in Conflict over Charter Schools.”
Donald Rosdil, Northern Virginia Community College: “The Role
of Comparative Case Analysis in Explaining Progressive Policy
Outcomes in U.S. Cities.”
Discussant: Curtis Nichols, University of Texas-Austin
Repression and Protest in Non-Democratic Regimes
Chair: Piero Stanig, Columbia University
Participants:
Mirjam Kunkler, Princeton University: “Protest and Repression
Cycles in Reformist Iran 1997–2001.”
Emmanuel Teitelbaum, George Washington University: “The
Effects of Labor Standards on Export Performance in Low- and
Middle-Income Countries.”
Jillian M. Schwedler, University of Massachusetts-Amherst:
“Political Protest in Neo-Liberal Jordan.”
Holger Albrecht and Kevin Koehler, American University in Cairo:
“Success and Failure of Protest Movements Under Authoritar-
ianism: Evidence from Burma, Egypt and Kyrgyzstan.”
Discussant: Maria Inclan, Centro de Investigación y Docencia
Económicas (CIDE)
Challenges and Advances in Historically-Oriented Research
Chair: Eileen M. Doherty-Sil, University of Pennsylvania
Participants:
Hillel David Soifer, Princeton University: “Permissive and Causal
Conditions in Historical Causation: Windows of Opportunity
and Types of Critical Junctures.”
Sean L. Yom, Harvard University: “Theoretical Deduction or
Empirical Induction? Resolving Causal Tensions in Comparative-
Historical Research.”
Tonya Caprarola Giannoni, George Washington University: “Ap-
plying Historical Methods to Understanding the Evolution of
Property Rights When Land is Not Scarce.”
Ian S. Lustick, University of Pennsylvania: “Evolution and His-
torical Institutionalism: Tropes Without Theory.”
Laura J. Hatcher, Southern Illinois University: “Critical Junctures
and Legal Meaning: (Re)Constituting Property by Constituting
Wetlands.”

Refinements in Research Design: Cases, Concepts, Variables
Chair: Carolyn M. Warner, Arizona State University
Participants:
Thomas Pluemper, Eric Neumayer, and Vera Troeger, University of
Essex, United Kingdom: “Case Selection in Qualitative Re-
search.”
Carolyn M. Warner, Arizona State University: “The Possibility
Principle and the Methodology of Comparative Case Studies of
Corruption.”
Etel L. Solingen, University of California-Irvine: “Theory and
Method in the Study of Nuclear Proliferation.”
Stefanie Walter and Dirk Leuffen, Harvard University: “Improving
Measurement in Qualitative Social Science Research.”
Discussant: Hillel David Soifer, Princeton University
Everyday Politics in Developing Countries:
Qualitative Approaches
Chair: Staffan I. Lindberg, University of Florida
Participants:
Maren Milligan, University of Maryland: “Navigating Shifting
Front Lines: A Relational Approach to Gender and Political
Ethnography in Religiously Divided Societies, A Comparison of
Nigeria and Lebanon.”
Sybille Ngo Nyeck, University of California-Los Angeles: “The
‘Dependent-Variable Problem’ of the Colonial State: Conceptual
Stretching and Discontent in Development Studies.”
Rodrigo Velazquez, University of Texas-Austin: “Democracy’s
Impact on Bureaucratic-Legislative Relations: Theoretical
Expectations and Mexican Realities.”
Jeremy Matthew Menchik, University of Wisconsin: “Fatwas as
Data: Uncovering Historical Change in Islamic Institutions.”
Discussant: Staffan I. Lindberg, University of Florida
Meaning, Discourse and Agency in Political Life
Chair: Chris Mantzavinos, Witten/Herdecke University, Germany
Participants:
Chris Mantzavinos, Witten/Herdecke University, Germany: “How
to Explain ‘Meaningful’ Actions.”
Brendan Jerome Hogan, New York University: “Imagination,
Political Science, and Agency.”
J. Mitchell Pickerill, Washington State University: “Designing
Conversational Interviews for Phronetic and Causal Analyses:
The Constitution in the Everyday Lives of Ordinary Americans.”
Eric M. Blanchard, University of Southern California: “What’s at
Stake in US-China Relations? Interpretation, Discourse Analysis
and the ‘Responsible Stakeholder’ Debate.”
Discussant: Jeremy Matthew Menchik, University of Wisconsin
Qualitative Approaches to Studying the
Emergence and Practice of Democracy
Chair: Djamel Mermat, University of Lille 2, France
Participants:
Mohamed Charfi, University of Geneva, Switzerland: “Democrati-
sation Processes in Arab Countries: A Fuzzy Set Analysis.”
Djamel Mermat and Monia Chaabane, University of Lille 2, France:
“Citizens in Front of their Screens: The Joint Influence of Al
Jazeera and TF1 on the Electoral Choice of French Voters of
Maghreb Origin.”
Ivo Lima Veiga, University College, London: “The Role of Coali-
tions in the Spanish and the Portuguese Transition to Democracy
1974-1978.”
Qualitative & Multi-Method Research, Spring 2009
Ryan Geoffery Baird, University of Arizona: “Governance Infra-
structure and High Quality Democracy: A Theoretically Moti-
vated Concept Construction and Necessary Condition Analy
sis.”
Discussant: Patricia J. Woods, University of Florida
Constructivism and Traditional IR Theory:
Pluralism, Conflict, or Eclecticism?
Chair: J. Samuel Barkin, University of Florida
Participants:
Craig A. Parsons, University of Oregon: “What Is Distinctive
about Constructivism?”
J. Samuel Barkin, University of Florida: “Realism, Constructivism,
and International Relations Theory.”
Cynthia S. Kaplan, University of California-Santa Barbara: “Test-
ing Constructivist Identity: Developing Empirical Indicators
from In-Depth Interviews.”
Jérémie Cornut, Université du Québec à Montréal: “Pluralism in
IR Theory: An Eclectic Study of Diplomatic Apologies and
Regrets.”
Discussant: Craig A. Parsons, University of Oregon
Constructing Cross-National Datasets:
Challenges and Lessons
Chair: Andreas Schedler, Centro de Investigación y Docencia
Económicas (CIDE)
Participants: Amy R. Poteete, Concordia University; Ronald A.
Francisco, University of Kansas; Monty G. Marshall, George
Mason University; Amy G. Mazur, Washington State Univer-
sity; Wolfgang Merkel, Wissenschaftszentrum Berlin für Sozial-
forschung (WZB), Germany
Discussant: Jose Antonio Cheibub, University of Illinois-Urbana-
Champaign
Research Design, Methods, and Theory-Building
in Comparative Judicial Politics
Chair: Jeffrey Staton, Emory University
Participants:
Matthew C. Ingram, University of New Mexico: “Bridging
Theory, Building Courts: Crossing Subfield Boundaries to
Clarify Causation in Judicial Politics.”
Beth Neitzel, University of California-Irvine: “Unfinished Busi-
ness: Examining the Meaning and Implications of Political
Fragmentation for Judicial Institutions and Behavior.”
Juan Rebolledo and Frances Rosenbluth, Yale University: “Mea-
suring the Rule of Law.”
Druscilla L. Scribner, University of Wisconsin-Oshkosh: “Select-
ing and Collecting Data in Comparative Judicial Politics.”
Discussant: Jeffrey Staton, Emory University
51
Qualitative and Multi-Method Research Nonprofit Org.
Department of Political Science U.S. Postage
315 Social Sciences Building PAID
P.O. Box 210027 TUCSON, AZ
University of Arizona Permit No. 271
Tucson, AZ 85721-0027
USA

Qualitative and Multi-Method Research is edited by Gary Goertz (tel: 520-621-1346, fax: 520-621-5051, email: ggoertz@u.arizona.edu). The
assistant editor is Joshua C. Yesnowitz (email: jcyesnow@bu.edu). Published with financial assistance from the Consortium for
Qualitative Research Methods (CQRM). Opinions do not represent the official position of CQRM. After a one-year lag, past issues will
be available to the general public online, free of charge, at http://www.maxwell.syr.edu/moynihan/programs/cqrm/section.html.
Annual section dues are $8.00. You may join the section online (http://www.apsanet.org) or by phone (202-483-2512). Changes of
address take place automatically when members change their addresses with APSA. Please do not send change-of-address information
to the newsletter.

52